Australia’s NationalScience Agency Characterising groundwater resources ofthe Gilbert River Formation, CamoowealDolostone and Thorntonia Limestone inthe Southern Gulf catchments, Queensland and Northern Territory A technical report fromthe CSIRO Southern GulfWater ResourceAssessment for theNationalWater Grid Matthias Raiber1,Andrew R Taylor1, Margaux Dupuy1, Stacey Priestley1,Karen Barry1, RussellCrosbie1,AnthonyKnapton2, Geoff Hodgson1 1 CSIRO, 2CloudGMS A blue and white cloud logo Description automatically generated ISBN 978-1-4863-2078-3 (online) A blue and white cloud logo Description automatically generated ISBN 978-1-4863-2077-6 (print) Our research direction Citation Raiber M, Taylor AR, Dupuy M, Priestley S, Barry K, Crosbie R, Knapton A and Hodgson G (2024) Characterising groundwater resources of the Gilbert River Formation, Camooweal Dolostone and Thorntonia Limestone in the Southern Gulf catchments, Queensland and Northern Territory. A technical report from the CSIRO Southern Gulf Water Resource Assessment for the National Water Grid. CSIRO, Australia. Copyright © Commonwealth Scientific and Industrial Research Organisation 2024. To the extent permitted by law, all rights are reserved and no part of this publication covered by copyright may be reproduced or copied in any form or by any means except with the written permission of CSIRO. Important disclaimer CSIRO advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it. CSIRO is committed to providing web accessible content wherever possible. If you are having difficulties with accessing this document please contact Email CSIRO Enquiries . CSIRO Southern Gulf Water Resource Assessment acknowledgements This report was funded through the National Water Grid’s Science Program, which sits within the Australian Government’s Department of Climate Change, Energy, the Environment and Water. Aspects of the Assessment have been undertaken in conjunction with the Northern Territory and Queensland governments. The Assessment was guided by two committees: i. The Governance Committee: CRC for Northern Australia/James Cook University; CSIRO; National Water Grid (Department of Climate Change, Energy, the Environment and Water); Northern Land Council; NT Department of Environment, Parks and Water Security; NT Department of Industry, Tourism and Trade; Office of Northern Australia; Queensland Department of Agriculture and Fisheries; Queensland Department of Regional Development, Manufacturing and Water ii. The Southern Gulf catchments Steering Committee: Amateur Fishermen’s Association of the NT; Austral Fisheries; Burketown Shire; Carpentaria Land Council Aboriginal Corporation; Health and Wellbeing Queensland; National Water Grid (Department of Climate Change, Energy, the Environment and Water); Northern Prawn Fisheries; Queensland Department of Agriculture and Fisheries; NT Department of Environment, Parks and Water Security; NT Department of Industry, Tourism and Trade; Office of Northern Australia; Queensland Department of Regional Development, Manufacturing and Water; Southern Gulf NRM Responsibility for the Assessment’s content lies with CSIRO. The Assessment’s committees did not have an opportunity to review the Assessment results or outputs prior to their release. This report was reviewed by Dr Jodi Pritchard (CSIRO) and Dr David Rassam (CSIRO). Acknowledgement of Country CSIRO acknowledges the Traditional Owners of the lands, seas and waters, of the area that we live and work on across Australia. We acknowledge their continuing connection to their culture and pay our respects to their elders past and present. Photo Lawn Hill Gorge. Source: Russell Crosbie, CSIRO Director’s foreword Sustainable development and regional economic prosperity are priorities for the Australian, Queensland and Northern Territory (NT) governments. However, more comprehensive information on land and water resources across northern Australia is required to complement local information held by Indigenous Peoples and other landholders. Knowledge of the scale, nature, location and distribution of likely environmental, social, cultural and economic opportunities and the risks of any proposed developments is critical to sustainable development. Especially where resource use is contested, this knowledge informs the consultation and planning that underpin the resource security required to unlock investment, while at the same time protecting the environment and cultural values. In 2021, the Australian Government commissioned CSIRO to complete the Southern Gulf Water Resource Assessment. In response, CSIRO accessed expertise and collaborations from across Australia to generate data and provide insight to support consideration of the use of land and water resources in the Southern Gulf catchments. The Assessment focuses mainly on the potential for agricultural development, and the opportunities and constraints that development could experience. It also considers climate change impacts and a range of future development pathways without being prescriptive of what they might be. The detailed information provided on land and water resources, their potential uses and the consequences of those uses are carefully designed to be relevant to a wide range of regional-scale planning considerations by Indigenous Peoples, landholders, citizens, investors, local government, and the Australian, Queensland and NT governments. By fostering shared understanding of the opportunities and the risks among this wide array of stakeholders and decision makers, better informed conversations about future options will be possible. Importantly, the Assessment does not recommend one development over another, nor assume any particular development pathway, nor even assume that water resource development will occur. It provides a range of possibilities and the information required to interpret them (including risks that may attend any opportunities), consistent with regional values and aspirations. All data and reports produced by the Assessment will be publicly available. Chris Chilcott Project Director A close-up of a black text Description automatically generated The Southern Gulf Water Resource Assessment Team Project Director Chris Chilcott Project Leaders Cuan Petheram, Ian Watson Project Support Caroline Bruce, Seonaid Philip Communications Emily Brown, Chanel Koeleman, Jo Ashley, Nathan Dyer Activities Agriculture and socio- economics Tony Webster, Caroline Bruce, Kaylene Camuti1, Matt Curnock, Jenny Hayward, Simon Irvin, Shokhrukh Jalilov, Diane Jarvis1, Adam Liedloff, Stephen McFallan, Yvette Oliver, Di Prestwidge2, Tiemen Rhebergen, Robert Speed3, Chris Stokes, Thomas Vanderbyl3, John Virtue4 Climate David McJannet, Lynn Seo Ecology Danial Stratford, Rik Buckworth, Pascal Castellazzi, Bayley Costin, Roy Aijun Deng, Ruan Gannon, Steve Gao, Sophie Gilbey, Rob Kenyon, Shelly Lachish, Simon Linke, Heather McGinness, Linda Merrin, Katie Motson5, Rocio Ponce Reyes, Jodie Pritchard, Nathan Waltham5 Groundwater hydrology Andrew R. Taylor, Karen Barry, Russell Crosbie, Margaux Dupuy, Geoff Hodgson, Anthony Knapton6, Stacey Priestley, Matthias Raiber Indigenous water values, rights, interests and development goals Pethie Lyons, Marcus Barber, Peta Braedon, Petina Pert Land suitability Ian Watson, Jenet Austin, Bart Edmeades7, Linda Gregory, Ben Harms10, Jason Hill7, Jeremy Manders10, Gordon McLachlan, Seonaid Philip, Ross Searle, Uta Stockmann, Evan Thomas10, Mark Thomas, Francis Wait7, Peter Zund Surface water hydrology Justin Hughes, Matt Gibbs, Fazlul Karim, Julien Lerat, Steve Marvanek, Cherry Mateo, Catherine Ticehurst, Biao Wang Surface water storage Cuan Petheram, Giulio Altamura8, Fred Baynes9, Jamie Campbell11, Lachlan Cherry11, Kev Devlin4, Nick Hombsch8, Peter Hyde8, Lee Rogers, Ang Yang Note: Assessment team as at September, 2024. All contributors are affiliated with CSIRO unless indicated otherwise. Activity Leaders are underlined. 1James Cook University; 2DBP Consulting; 3Badu Advisory Pty Ltd; 4Independent contractor; 5 Centre for Tropical Water and Aquatic Ecosystem Research. James Cook University; 6CloudGMS; 7NT Department of Environment, Parks and Water Security; 8Rider Levett Bucknall; 9Baynes Geologic; 10QG Department of Environment, Science and Innovation; 11Entura Shortened forms For more information on this table please contact CSIRO on enquiries@csiro.au For more information on this table please contact CSIRO on enquiries@csiro.au Units UNIT DESCRIPTION % percent ‰ per mille cm centimetre g gram GL gigalitre ha hectare kg kilogram km kilometre L litre m metre Ma Megaannum, or million years ago meq milliequivalent mg milligram ML megalitres mm millimetre mmol millimole mS millisiemen pMC percent modern carbon s second y year μS microsiemen Preface Sustainable development and regional economic prosperity are priorities for the Australian, NT and Queensland governments. In the Queensland Water Strategy, for example, the Queensland Government (2023) looks to enable regional economic prosperity through a vision that states ‘Sustainable and secure water resources are central to Queensland’s economic transformation and the legacy we pass on to future generations.’ Acknowledging the need for continued research, the NT Government (2023) announced a Territory Water Plan priority action to accelerate the existing water science program ‘to support best practice water resource management and sustainable development.’ Governments are actively seeking to diversify regional economies, considering a range of factors, including Australia’s energy transformation. The Queensland Government’s economic diversification strategy for North West Queensland (Department of State Development, Manufacturing, Infrastructure and Planning, 2019) includes mining and mineral processing; beef cattle production, cropping and commercial fishing; tourism with an outback focus; and small business, supply chains and emerging industry sectors. In its 2024–25 Budget, the Australian Government announced large investment in renewable hydrogen, low-carbon liquid fuels, critical minerals processing and clean energy processing (Budget Strategy and Outlook, 2024). This includes investing in regions that have ‘traditionally powered Australia’ – as the North West Minerals Province, situated mostly within the Southern Gulf catchments, has done. For very remote areas like the Southern Gulf catchments (Preface Figure 1-1), the land, water and other environmental resources or assets will be key in determining how sustainable regional development might occur. Primary questions in any consideration of sustainable regional development relate to the nature and the scale of opportunities, and their risks. How people perceive those risks is critical, especially in the context of areas such as the Southern Gulf catchments, where approximately 27% of the population is Indigenous (compared to 3.2% for Australia as a whole) and where many Indigenous Peoples still live on the same lands they have inhabited for tens of thousands of years. About 12% of the Southern Gulf catchments are owned by Indigenous Peoples as inalienable freehold. Access to reliable information about resources enables informed discussion and good decision making. Such information includes the amount and type of a resource or asset, where it is found (including in relation to complementary resources), what commercial uses it might have, how the resource changes within a year and across years, the underlying socio-economic context and the possible impacts of development. Most of northern Australia’s land and water resources have not been mapped in sufficient detail to provide the level of information required for reliable resource allocation, to mitigate investment or environmental risks, or to build policy settings that can support good judgments. The Southern Gulf Water Resource Assessment aims to partly address this gap by providing data to better inform decisions on private investment and government expenditure, to account for intersections between existing and potential resource users, and to ensure that net development benefits are maximised. Preface Figure 1-1 Map of Australia showing Assessment area (Southern Gulf catchments) and other recent CSIRO Assessments FGARA = Flinders and Gilbert Agricultural Resource Assessment; NAWRA = Northern Australia Water Resource Assessment. The Assessment differs somewhat from many resource assessments in that it considers a wide range of resources or assets, rather than being a single mapping exercises of, say, soils. It provides a lot of contextual information about the socio-economic profile of the catchments, and the economic possibilities and environmental impacts of development. Further, it considers many of the different resource and asset types in an integrated way, rather than separately. The Assessment has agricultural developments as its primary focus, but it also considers opportunities for and intersections between other types of water-dependent development. For example, the Assessment explores the nature, scale, location and impacts of developments relating to industrial, urban and aquaculture development, in relevant locations. The outcome of no change in land use or water resource development is also valid. The Assessment was designed to inform consideration of development, not to enable any particular development to occur. As such, the Assessment informs – but does not seek to replace – existing planning, regulatory or approval processes. Importantly, the Assessment does not assume a given policy or regulatory environment. Policy and regulations can change, so this flexibility enables the results to be applied to the widest range of uses for the longest possible time frame. It was not the intention of – and nor was it possible for – the Assessment to generate new information on all topics related to water and irrigation development in northern Australia. Topics For more information on this figure please contact CSIRO on enquiries@csiro.au not directly examined in the Assessment are discussed with reference to and in the context of the existing literature. CSIRO has strong organisational commitments to Indigenous reconciliation and to conducting ethical research with the free, prior and informed consent of human participants. The Assessment allocated significant time to consulting with Indigenous representative organisations and Traditional Owner groups from the catchments to aid their understanding and potential engagement with its requirements. The Assessment did not conduct significant fieldwork without the consent of Traditional Owners. CSIRO met the requirement to create new scientific knowledge about the catchments (e.g. on land suitability) by synthesising new material from existing information, complemented by remotely sensed data and numerical modelling. Functionally, the Assessment adopted an activities-based approach (reflected in the content and structure of the outputs and products), comprising activity groups, each contributing its part to create a cohesive picture of regional development opportunities, costs and benefits, but also risks. Preface Figure 1-2 illustrates the high-level links between the activities and the general flow of information in the Assessment. Preface Figure 1-2 Schematic of the high-level linkages between the eight activity groups and the general flow of information in the Assessment Assessment reporting structure Development opportunities and their impacts are frequently highly interdependent and, consequently, so is the research undertaken through this Assessment. While each report may be read as a stand-alone document, the suite of reports for each Assessment most reliably informs discussion and decisions concerning regional development when read as a whole. For more information on this figure please contact CSIRO on enquiries@csiro.au The Assessment has produced a series of cascading reports and information products: • Technical reports present scientific work with sufficient detail for technical and scientific experts to reproduce the work. Each of the activities (Preface Figure 1-2) has one or more corresponding technical reports. • A catchment report, which synthesises key material from the technical reports, providing well- informed (but not necessarily scientifically trained) users with the information required to inform decisions about the opportunities, costs and benefits, but also risks, associated with irrigated agriculture and other development options. • A summary report provides a shorter summary and narrative for a general public audience in plain English. • A summary fact sheet provides key findings for a general public audience in the shortest possible format. The Assessment has also developed online information products to enable users to better access information that is not readily available in print format. All of these reports, information tools and data products are available online at https://www.csiro.au/southerngulf. The webpages give users access to a communications suite including fact sheets, multimedia content, FAQs, reports and links to related sites, particularly about other research in northern Australia. Executive summary The catchments of the Southern Gulf rivers, that is Settlement Creek, Gregory—Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups1, in the NT and Queensland, encompass an area of approximately 108,200 km2 within the wet-dry tropics of northern Australia. The Southern Gulf catchments can generally be split into the uplands (highlands and plateaux) and the coastal plains. The uplands area in the south and west of the catchments reaches a maximum of about 620 m above mean sea level and hosts the headwaters for major streams (e.g. Lawn Hill Creek and the O’Shannassy and Gregory rivers) in the catchments. To the east of the Uplands are the Carpentaria Plains, comprising a series of alluvial and coastal plains, pediments and remnant plateaux. 1 Only those islands greater than 1000 ha are mapped. The Southern Gulf catchments overlie six major geological sedimentary basins. These are from oldest to youngest: (i) the late Paleoproterozoic to early Mesoproterozoic (~1780–1400 Ma) McArthur Basin, which underlies a small portion of the catchment north-west (mostly below Settlement Creek), (ii) the Paleoproterozoic–Mesoproterozoic (1670 to 1575 Ma) Isa Superbasin, (iii) the Mesoproterozoic (1483 to 1266 Ma) South Nicholson Basin, (iv) the Neoproterozoic to late Palaeozoic (850 to 350 Ma) Georgina Basin, which overlies the South Nicholson Basin in the south- west of the catchment, (v) the Jurassic to Cretaceous geological Carpentaria Basin, which is the northern-most sub-basin of the Great Artesian Basin (GAB) and underlies most of the east and north-eastern parts of the Southern Gulf catchments and (vi) the Cenozoic Karumba Basin, which covers most of the Carpentaria Basin in the north-eastern part of the Southern Gulf catchments. Overlying the Karumba Basin are the youngest sediments in the catchments, the alluvial sands, silts and gravels associated with the beds, channels and floodplains of the catchment’s rivers and creeks. Within the Southern Gulf catchments, the availability and quality of groundwater resources are heavily influenced by the physical characteristics of the strata within the sedimentary basins, fractured rock provinces and alluvial sediments. Different aquifer types are hosted in each of the geological provinces across the catchment, including: (i) local to intermediate-scale fractured and weathered rocks and porous sandstones; (ii) fractured, fissured and karstic carbonate rocks of the Georgina Basin; (iii) extensive variably productive porous sedimentary sandstones of the Carpentaria Basin; (iv) porous sandstones of the Karumba Basin; and (v) surficial unconsolidated to consolidated alluvial sands and gravels. This groundwater study, which formed the groundwater hydrology component of the Southern Gulf Water Resource Assessment, aimed at identifying the most promising opportunities for future groundwater resource development. It included: (i) a literature and data review of all previous hydrogeological investigations in the catchments, (ii) a regional-scale desktop data collation and analyses, including digitising data contained in hand-written and typed drilling records, and evaluating groundwater levels, groundwater salinity and bore yields, (iii) a regional-scale recharge modelling assessment of all aquifers, and (iv) regional-scale mapping of potential groundwater discharge areas. Collectively, results of the literature and data review were used to identify the most promising intermediate- to regional-scale aquifers for undertaking more detailed, desktop and modelling investigations. The key aim of these more detailed investigations of the most promising aquifers (a major focus of this study) was to obtain detailed information to support future decisions around the planning, investment and management of key groundwater resources in the catchments. The literature review provided valuable insight into all aquifers and their current knowledge gaps. The regional-scale desktop and modelling assessment served as useful screening tool for identifying spatial patterns in important groundwater attributes, including: (i) the spatial extent of aquifers, (ii) ranges for groundwater levels, groundwater salinity and aquifer hydraulic properties, and (iii) ranges in groundwater recharge and mapping of potential groundwater discharge features across the catchment. It also helped to provide the baseline datasets used in detailed desktop and modelling analyses (e.g. the recharge estimation) and numerical modelling. A spatial evaluation of key groundwater attributes highlighted the geographic regions beneath the catchment that contained aquifers hosting either local-scale (1 to 10 km), intermediate-scale (10 to 50 km) or regional-scale (>50 to >100 km) groundwater flow systems. Localised aquifers included the Cenozoic alluvial aquifers in the mid- and lower reaches of the Leichhardt, Nicholson and Gregory rivers, and Settlement Creek and their tributaries. Localised aquifers are also hosted within the various Proterozoic igneous, metasedimentary and metamorphic rocks, all of which are currently poorly characterised. The available data suggest that these aquifers contain good-quality groundwater (with less than 500 mg/L total dissolved solids (TDS)) in places, but are brackish (with a TDS of more than7500 mg/L) elsewhere. Yields are mostly less than 2 L/second, based on sparse data. These aquifers provide an important small-scale source of water for stock, domestic and occasional mining use, but limited potential for groundwater-based irrigation The regional aquifers that were identified include: (i) the regional-scale Cambrian Limestone Aquifer (CLA) hosted within the Camooweal Dolostone and Thorntonia Limestone (which are part of the geological Georgina Basin) and (ii) the regional-scale Gilbert River Aquifer (GRA) hosted in the Gilbert River Formation within the geological Carpentaria Basin (the northern-most sub-basin of the GAB). The range of the salinity of the regional-scale aquifers within the Southern Gulf catchments was generally less (<500 to 3000 mg/L) than the localised aquifers. Indicative bore yields in the CLA were low (2–10 L/second), based on sparse aquifer testing. However, pumping tests indicate that bore yields of >20 L/second are achievable where appropriately constructed production bores have been installed in highly karstic parts (large, interconnected caves, caverns and fractures) of the aquifer. Aquifer testing in the GRA is sparse, but indicative bore yields are often high (10–20 L/second) with some exceeding 20 L/second. These two aquifers (CLA and GRA) were identified as the most promising for future groundwater resource development. Cambrian Limestone Aquifer Recharge rates to the CLA (comprised of the Camooweal Dolostone and Thorntonia Limestone), estimated by the upscaled chloride mass balance (CMB) technique, indicate spatially variable recharge rates ranging between approximately 3 and 30 mm/year. Recharge to the CLA occurs as broad diffuse recharge via vertical leakage across almost the entire spatial extent of the aquifer within the Southern Gulf catchments, but is likely most prominent in areas where the carbonate aquifers are not covered by Cenozoic sediments and/or the siltstones of the Cambrian Wonarah Formation. Localised recharge occurs: (i) via direct infiltration of rainfall and streamflow via sinkholes in and near the aquifer outcrop, or (ii) where surficial features (sinkholes, ephemeral streams, or waterholes) are incised through thinner parts of the overlying Wonarah Formation or Cenozoic strata along the aquifer margins. Mapping the hydraulic head across the CLA has further confirmed the complexity of the groundwater flow dynamics within the aquifer. Although previous studies suggested that discharge to Lawn Hill Creek and the Gregory River is sourced from as far west as the Alexandria– Wonarah High, the groundwater head map constructed as part of the groundwater model development indicated that another structural high east of this, influences a regional groundwater divide separating groundwater flows to the east and south. This agrees with findings of a recent study by Geoscience Australia on the hydrogeology and groundwater systems of the South Nicholson and Georgina basins, which named the east-flowing groundwater system of the CLA in the Georgina Basin part of the Southern Gulf catchments as the Lawn Hill Creek groundwater flow system. This term was adopted for the following discussion on discharge processes. Although the CLA is classified as a regional flow system, localised recharge and discharge in the fractured and karstic aquifer outcrop are also important processes, particularly for supporting spring flow and baseflow to parts of the regional groundwater discharge zone in the eastern part of the Lawn Hill Creek groundwater flow system. A two-dimensional numerical groundwater flow model has been developed using Finite Element subsurface FLOW system (FEFLOW) to examine the groundwater balance of the CLA within the Undilla Sub-basin of the Georgina Basin. The Undilla Sub-basin encompasses the Lawn Hill Creek groundwater flow system that provides baseflow to Lawn Hill Creek and Gregory and O’Shannassy rivers. The conceptualisation of the groundwater flow system indicated that there is a localised system discharging to springs well above the stream level and a regional groundwater system discharging to lower springs and through the bed of the river. Depth to groundwater maps, streamflow gauging time series data and the analysis of actual evapotranspiration (AET) were used to identify areas where dry-season AET was higher than expected. This indicated that discharge likely occurs via a combination of: (i) lateral outflow to streams where they are incised in the aquifer outcrop (Gregory and O’Shannassy rivers and Lawn Hill Creek), (ii) localised spring discharge, (iii) transpiration from riparian, spring- fed and other phreatophytic vegetation where the watertable for the Cambrian dolostone and limestone is shallow (i.e. ≤5 m below ground level (BGL)) in the south-eastern part of Lawn Hill Creek groundwater flow system and (iv) limited groundwater extraction. Most mapped springs within the headwaters of Lawn Hill Creek and the Gregory and O’Shannassy rivers are located outside the geological Georgina Basin and within the outcrop of the Constance Sandstone, which is part of the South Nicholson Basin. Geological structures and inter-basin flow from the Georgina Basin to the Constance Sandstone were hypothesised as potential discharge pathways to these springs in a recent study by Geoscience Australia on the hydrogeology and groundwater systems of the South Nicholson and Georgina basins. This hypothesis was based on the interpretation of airborne electromagnetic (AEM) survey data. However, the study also acknowledged the lack of hydrochemistry and isotope data and higher resolution AEM surveys as key data gaps. Based on the available data, other mechanisms, such as localised recharge and discharge processes within the Constance Sandstone, are also plausible as discharge pathways to these mapped springs. Gilbert River Aquifer The geological Carpentaria Basin is a portion of the GAB that underlies most of the eastern parts of the Southern Gulf catchments. The GAB is comprised of Jurassic- to Cretaceous-aged interbedded sandstones, mudstones and siltstones, which become deeper and thicker towards the north of the study area. The most productive aquifer of the Carpentaria Basin within the Southern Gulf catchments is the Gilbert River Formation (also referred to in this document as the GRA). Mineralogical analyses conducted during this study on cores from Dobbyn-1 exploration wells at the southern edge of the Southern Gulf catchments highlighted that the GRA is likely to have significant permeability and is thus a productive aquifer at this location. Overlying the GRA are the Lower to Upper Cretaceous rock sequences of the Rolling Downs Group. The Rolling Downs Group comprises the Normanton, Toolebuc and Wallumbilla formations and the Allaru Mudstone. Mineralogical analyses conducted during this assessment on cores from Dobbyn-1 exploration wells at the southern edge of the Southern Gulf catchments have indicated that the rock sequences of the Rolling Downs Group are likely to represent regional aquitards or leaky aquitards at this location. Only the Normanton Formation is known in places to host a partial aquifer of limited productivity and variable water quality. Recharge to the GRA and the Rolling Downs Group in the Southern Gulf catchments remains a key knowledge gap. Regional recharge estimates on surficial Cretaceous sediments outside the Southern Gulf catchments using the CMB method are estimated to be 26 mm/year (ranging between 15 and 46 mm/year). However, the GRA and Normanton Formation do not outcrop in the Southern Gulf catchments, and consequently there is no direct recharge to these aquifers here. Recharge to the GRA and the Rolling Downs Group is, therefore, likely sourced from throughflow outside the Southern Gulf catchments and/or from possible vertical leakage from overlying Cenozoic sediments at the western margin of both aquifers in the Southern Gulf catchments. The GRA is shallowest (<150 mBGL) at its western basin margin in the mid-reaches of the study area and increases in depth in a north-east direction towards the coast (>400 mBGL). Groundwater flow in the GRA is towards the coast, and the GRA extends offshore at great depths and increasing thicknesses. The GRA is sub-artesian near the basin margin, with artesian conditions (groundwater under pressure and naturally rising to the surface without pumping) being established approximately 20 to 50 km from its western margin and likely extending into the Gulf of Carpentaria. It is unknown whether the GRA contributes groundwater discharge to springs and seeps within the study area. However, given the depth of the aquifer and the thickness of the overlying low-permeability rocks of the Rolling Downs Group, discharge to springs in the Southern Gulf catchments is considered unlikely. Furthermore, no springs linked to the GRA or partial aquifers hosted in the overlying Rolling Downs Group are mapped within the extent of the Carpentaria Basin in the Southern Gulf catchments. This indicates that groundwater discharge from the GAB is most likely offshore to the Gulf of Carpentaria. However, the exact locations and volumes of discharge are unknown, as is the potential for upward leakage, hence they are currently deemed a knowledge gap. The mean groundwater salinity in the GRA within the Southern Gulf catchments is 1915 μS/cm (median 1854 μS/cm) in the GRA within the Southern Gulf catchments. Salinities are lowest near the basin margin (mostly below 1200 μS/cm) and increase along inferred flow paths towards the Gulf of Carpentaria. Groundwater development opportunities The hydrogeological units of the Southern Gulf catchments contain a variety of local- and regional- scale aquifers that host localised to regional-scale groundwater flow systems. The study identified five hydrogeological units hosting aquifers that may have potential for future groundwater resource development in the Southern Gulf catchments: • Cambrian limestone and dolostone • Gilbert River Formation • Cenozoic alluvium • Proterozoic igneous rocks • Proterozoic metasedimentary and metamorphic rocks. The regional-scale CLA hosted in the Cambrian limestone and dolostone and GRA hosted in the Gilbert River Formation, provide the greatest potential opportunities for future groundwater resource development across the Southern Gulf catchments. They are present in the subsurface across large areas, collectively occurring beneath approximately 45% of the catchment. Given their large spatial extent, they also underlie and coincide frequently with larger areas of soil suitable for irrigated agriculture. They contain areas of mostly low-salinity water (<1000 mg/L TDS) and can yield water at a sufficient rate to support irrigation development (>10 L/second). These aquifers also store larger volumes of groundwater (gigalitres to teralitres) compared to local-scale aquifers with storage and discharge characteristics often less affected by short-term (annual) variations in recharge rates (driven by inter-annual variability in rainfall). Furthermore, their larger spatial extent provides greater opportunities for groundwater resource development away from existing groundwater users and groundwater-dependent ecosystems (GDEs) at the land surface, such as springs, spring-fed vegetation, and surface water, which can be ecologically and culturally significant. The most promising regional-scale aquifer is the CLA. The CLA is typically tens of metres thick and mostly hosts good-quality water (<1000 mg/L TDS). Indicative bore yields may be low (2– 10 L/second) due to sparse aquifer testing, but where appropriately constructed production bores have been installed in highly karstic parts (large, interconnected caves, caverns and fractures) of the aquifer, pumping tests indicate that bore yields of >20 L/second are achievable. It has the potential to support a few small (1–3 GL/year)- to intermediate-scale (3–5 GL/year) developments. The greatest opportunities exist along the south-western part of the Southern Gulf catchments boundary where the CLA outcrops. Opportunities are limited where the aquifer co-exists by the upper reaches of prescribed watercourses (Lawn Hill Creek, and the Gregory and O'Shannassy rivers that receive groundwater discharge from the CLA. The GRA is a potentially promising regional-scale sandstone aquifer. Aquifer testing is sparse, but indicative bore yields are often high (10–20 L/second) and have been found in places to exceed 20 L/second. The aquifer has variable water quality, with some parts being fresh (<1000 mg/L TDS) but in other places brackish (>2000 mg/L). It has the potential to support a few small-scale (1–2 GL/year) developments, with the greatest opportunities existing along the south-western margin of the GAB where the GRA is less than 500 m deep, and where the groundwater is fresh. In contrast, the local-scale aquifers in the Southern Gulf catchments, such as the fractured and weathered rock and alluvial aquifers, are relatively poorly characterised at present. They host local-scale groundwater systems that are highly variable in composition, salinity and yield. They also have a small and variable spatial extent with less storage compared with the larger aquifers, thus limiting groundwater resource development to localised opportunities such as for stock and domestic use, or as a conjunctive water resource (i.e. combined use of surface water, groundwater or rainwater). The most promising local-scale opportunities exist within the local-scale alluvial (sand and gravel) aquifers occurring in association with the stream channels, streambeds and floodplains of the mid- to lower reaches of the Leichhardt, Nicholson and Gregory rivers, and Settlement Creek and their tributaries. Data for these aquifers are sparse, but indicative bore yields range between 2 and 5 L/second, and indicative water-quality testing, though sparse, suggests they host good-quality water (<600 mg/L TDS) in places. Alluvial aquifers may have the potential for multiple small-scale (<0.5 GL/year) localised developments or as a conjunctive water resource where surface water is available. Local-scale opportunities also exist in local-scale fractured and weathered rock aquifers composed of a wide variety of volcanic rocks (basalt, breccia, rhyolite, tuff, agglomerate, or dolerite) and in fractured and weathered rock aquifers composed of a wide variety of metasedimentary (sandstone, siltstone, shale, conglomerate) and metamorphic (gneiss and schist) rocks. They are only likely to have the potential for small-scale (<0.5 GL/year and <0.25 GL/year, respectively) localised developments (i.e. mostly suited to stock and domestic water supplies) where fracturing and weathering is high. Constraints on groundwater development opportunities The mean groundwater salinities within some parts of the GRA are too high for several field crops, pastures, fruits and vegetables grown in loam or clay soils, such as rice, avocado or potatoes, but would be sufficient for almost all of these in sandy soils. The GRA increases in depth in a north-east direction towards the coast (>400 mBGL), with water becoming increasingly brackish (>2000 mg/L), thus presenting economic challenges for groundwater infrastructure. Opportunities are limited near ecologically and culturally important springs and seeps (although at present it is unknown whether any springs and seeps linked to the GRA exist within the Southern Gulf catchments). Opportunities are also likely to be limited near existing licensed groundwater users (such as the communities of Burketown and Gununa), and where the GRA is prohibitively deep (>500 mBGL) and/or where the groundwater is brackish (>2000 mg/L TDS). For the CLA opportunities are likely to be limited where the aquifer co-occurs by the upper reaches of the prescribed watercourses (reaches of Lawn Hill Creek, and the Gregory and O'Shannassy rivers). This is where groundwater discharge from the CLA supports ecologically and culturally important springs and seeps that support streamflow in the Gregory and O’Shannassy rivers and Lawn Hill Creek in the Nicholson Groundwater Management Area (NGMA) of the Gregory River Subcatchment Area of the Gulf Water Plan. Additional constraints include: • installation of successful productive high-yielding (i.e. bore yields of >20 L/second) bores may require the drilling of multiple investigation holes to identify productive parts of the aquifer; this can create uncertainty in the cost of drilling programs • the need to consider crop selection and irrigation of fine-textured soils across parts of the GRA and CLA as some areas exhibit slightly brackish groundwater (i.e. 1000–3000 mg/L TDS), which can degrade crops, depending on their specific salt tolerance, as well as cause soil structure decline (such as dispersion) • the need to maintain water quality across the aquifer, given some parts of the GRA and CLA are brackish with increasing irrigated agriculture potentially leading to rising aquifer salinity over time • the need to better characterise other potential areas of groundwater–surface water interactions, as these are likely to be a key constraint on future groundwater resource development in certain locations for both aquifers • the need to better characterise the water balance, saturated thickness, groundwater levels, and hydraulic properties of aquifers, as the CLA and GRA are currently data sparse in places across the Southern Gulf catchments • Cenozoic alluvial aquifers in the mid- to lower reaches of the Nicholson, Gregory and Leichhardt rivers, and Settlement Creek and its tributaries, host local-scale groundwater systems. Uncertainties and knowledge gaps The regional desktop assessment of all available hydrogeological data proved useful as a screening tool for identifying potential groundwater resource development opportunities across the Southern Gulf catchments. The regional-scale Cambrian dolostone and limestone aquifer and the regional-scale Gilbert River Formation provide the greatest potential opportunities for future groundwater resource development across the Southern Gulf catchments. However, the nature and scale of these opportunities will ultimately depend upon community and government acceptance of impacts to GDEs and existing groundwater users, as well as approval of licenses to extract groundwater. Groundwater systems in the Southern Gulf catchments are generally poorly studied. Although a targeted field investigation program was planned in the Southern Gulf catchments, this could not be completed due to land access challenges. Knowledge gaps and opportunities to close these gaps have been identified for these regional-scale aquifers and for local-scale aquifers, such as the alluvium. Key knowledge gaps and uncertainties include: Cambrian Limestone Aquifer • Groundwater-level time series data can inform groundwater-level dynamics of the CLA and its interaction with surface water features in the headwaters of Lawn Hill Creek and the Gregory and O’Shannassy rivers. • Additional rainfall chloride and isotope measurements throughout the Southern Gulf catchments would reduce the uncertainties of chloride mass balance recharge estimates. • The simultaneous collection and analysis of multiple environmental tracers (e.g. using tritium (3H), 14C and noble gases) along inferred groundwater flow paths would enhance the understanding of recharge mechanisms, including preferential recharge through karst features, and the independent quantification of recharge processes and mean residence times for groundwater flow. • Developing a hydrochemical and isotopic fingerprinting framework could help in identifying sources of discharge to Lawn Hill Creek (Lawn Hill Gorge) and the Gregory and O’Shannassy rivers. Gilbert River Aquifer • Improved mapping of the Carpentaria Basin geometry at the basin margin could help in identifying the location and source of recharge to the GRA and partial aquifers hosted in the Rolling Downs Group, which is likely to be at least partially from outside the Southern Gulf catchments. • Mineralogical characterisation (XRD and XRF) conducted on additional cores would verify the thickness and location of transmissive zones, particularly in the Rolling Downs Group. • Multi-tracer sampling (e.g. using 3H, 14C, 36Cl and stable and radioactive noble gases) along inferred flow paths to determine mean residence times for groundwater flow in the GRA and the Rolling Downs Group could reduce uncertainties associated with conceptual hydrogeological models. Cenozoic alluvial aquifers Cenozoic alluvial aquifers in the mid- to lower reaches of the Nicholson, Gregory and Leichhardt rivers, and Settlement Creek and its tributaries, host local-scale groundwater systems. However, these alluvial aquifers within the Southern Gulf catchments are poorly characterised, due to a lack of groundwater bores and monitoring of bores with reliable lithological, water chemistry, and hydraulic data. Some knowledge gaps and opportunities to close them include: • Geophysical surveys (e.g. ground-based or airborne electromagnetic surveys) could provide valuable insights into the extent, thickness and internal architecture of alluvial aquifers as well as identifying localised connectivity to streams. • Targeted drilling of additional bores, and instrumentation of bores with data loggers, could inform watertable fluctuations, aquifer water quality and yields, and the connectivity of aquifers with surface waters. • The collection and analysis of environmental tracers in alluvial groundwaters could inform mean residence times for groundwater flow, groundwater–surface water connectivity, and groundwater recharge processes. Contents Director’s foreword .......................................................................................................................... i The Southern Gulf Water Resource Assessment Team .................................................................. ii Shortened forms .............................................................................................................................iii Units ............................................................................................................................... v Preface ............................................................................................................................... vi Executive summary .......................................................................................................................... x Part I Introduction and overview 1 1 Introduction ........................................................................................................................ 2 1.1 Overview ................................................................................................................ 2 1.2 Aims of the groundwater hydrology activity ......................................................... 3 1.3 Summary of previous hydrogeological investigations .......................................... 3 1.4 Current water demand and licensed entitlements ............................................... 5 1.5 Report overview and structure ............................................................................. 8 2 Study area ........................................................................................................................... 9 2.1 Physiography and demography ............................................................................. 9 2.2 Climate ................................................................................................................. 12 2.3 Geology ................................................................................................................ 14 2.4 Hydrogeology....................................................................................................... 24 2.5 Surface water hydrology ..................................................................................... 34 2.6 Water-dependent ecosystems ............................................................................ 36 Part II Methods 39 3 Regional desktop and modelling assessment of the Southern Gulf catchments ............. 40 3.1 Regional geological and hydrogeological desktop assessment .......................... 40 3.2 Mineralogical analysis of core samples ............................................................... 41 3.3 Groundwater bore data ....................................................................................... 42 3.4 Aquifer data ......................................................................................................... 44 3.5 Groundwater levels ............................................................................................. 46 3.6 Groundwater chemistry and salinity ................................................................... 47 3.7 Bore yield ............................................................................................................. 48 3.8 Recharge estimation ............................................................................................ 49 3.9 Identifying potential groundwater discharge areas using remote sensing ........ 60 4 Detailed, desktop and modelling investigations .............................................................. 64 4.1 Hydrogeological framework ................................................................................ 64 4.2 Groundwater recharge and flow ......................................................................... 65 4.3 Groundwater salinity and hydrochemistry by stratigraphic formation .............. 65 4.4 Numerical flow modelling ................................................................................... 66 Part III Results 71 5 Regional desktop and modelling assessment of the Southern Gulf catchments ............. 72 5.1 Aquifer data ......................................................................................................... 72 5.2 Groundwater levels ............................................................................................. 73 5.3 Groundwater salinity and hydrochemistry ......................................................... 75 5.4 Bore yield ............................................................................................................. 83 5.5 Recharge estimation ............................................................................................ 85 5.6 Identifying potential groundwater discharge areas using remote sensing ........ 94 6 Detailed desktop and modelling investigations ............................................................. 101 6.1 Hydrogeological framework .............................................................................. 101 6.2 Groundwater recharge and flow ....................................................................... 114 6.3 Groundwater salinity and hydrochemistry by stratigraphic formation ............ 126 6.4 Numerical flow modelling ................................................................................. 156 Part IV Discussion and conclusions 159 7 Discussion ....................................................................................................................... 160 7.1 Cambrian limestone and dolostone .................................................................. 160 7.2 Gilbert River Formation and Rolling Downs Group ........................................... 163 7.3 Potential opportunities for future groundwater resource development ......... 165 8 Summary and conclusions .............................................................................................. 171 8.1 Opportunities and constraints for future groundwater development ............. 171 8.2 Potential options for future work...................................................................... 172 References ........................................................................................................................... 175 Figures Preface Figure 1-1 Map of Australia showing Assessment area (Southern Gulf catchments) and other recent CSIRO Assessments ................................................................................................... vii Preface Figure 1-2 Schematic of the high-level linkages between the eight activity groups and the general flow of information in the Assessment ..................................................................... viii Figure 1-1 Spatial distribution of surface water and groundwater licensed entitlements across the Southern Gulf catchments ........................................................................................................ 7 Figure 2-1 Physiographic regions of the Southern Gulf catchments ............................................ 10 Figure 2-2 Historical rainfall, potential evaporation and rainfall deficit ...................................... 13 Figure 2-3 Major geological provinces of the Southern Gulf catchments .................................... 15 Figure 2-4 Full extent of the Georgina Basin and Carpentaria Sub-basin of the Great Artesian Basin. Inset map shows full extent of Great Artesian Basin ......................................................... 16 Figure 2-5 Geological cross-section (west to east) through the Southern Gulf catchments (Buchanan et al., 2020) ................................................................................................................. 17 Figure 2-6 Composite logs of Carrara 1 stratigraphic well (Geoscience Australia, 2023) ............ 19 Figure 2-7 Surface geology of the Southern Gulf catchments ...................................................... 21 Figure 2-8 Major structural elements of the extended Southern Gulf catchments ..................... 23 Figure 2-9 Simplified regional aquifer-types of the Southern Gulf catchments ........................... 25 Figure 2-10 Cross-section of the Carpentaria and Karumba basins in the Isa Geological and Bioregional Assessment (GBA) region (Buchanan et al., 2020) .................................................... 28 Figure 2-11 Simplified regional hydrogeology of the Southern Gulf catchments ........................ 29 Figure 2-12 Thickness of Cenozoic sediments in the Southern Gulf catchments ......................... 33 Figure 2-13 Streamflow observation data availability and median annual streamflow (50% exceedance) under Scenario A in the Southern Gulf catchments ................................................ 35 Figure 2-14 Water-dependent ecosystems in Southern Gulf catchments including terrestrial, aquatic and subsurface environments that require surface water inundation and/or access to groundwater ................................................................................................................................. 37 Figure 3-1 Queensland core library at the Exploration Data Centre, Zillmere, Qld .................... 42 Figure 3-2 Simplified regional geology of the Southern Gulf catchments .................................... 45 Figure 3-3 Dendrogram of groundwater chemistry data showing separation threshold for five clusters. Each vertical blue line corresponds to one groundwater chemistry record .................. 48 Figure 3-4 Schematic diagram of the methods used showing the calculation of the point recharge, the upscaling using regression kriging and the reporting of the results at the scale of the region and surface geology group .......................................................................................... 50 Figure 3-5 Region used for estimating recharge using the chloride mass balance method ........ 51 Figure 3-6 The (a) mean, (b) standard deviation and (c) skewness of chloride deposition across the study area (Wilkins et al., 2022a). The blue squares indicate points where chloride deposition has been measured ..................................................................................................... 53 Figure 3-7 Runoff coefficient across the study area ..................................................................... 54 Figure 3-8 Covariates used in upscaling (a) rainfall, (b) clay content of the soil, (c) Normalised Difference Vegetation Index (NDVI) and (d) simplified geology ................................................... 57 Figure 3-9 (a) Catchments used for baseflow analysis for rejection sampling and (b) additional internally draining catchments used with excess water for rejection sampling .......................... 59 Figure 3-10 Relationship used for determining the threshold October actual evapotranspiration (AET) for estimating potential discharge areas from aridity index ............................................... 62 Figure 4-1 Location of the Undilla Sub-basin groundwater flow model in relation to the Southern Gulf catchments and portions of the model that coincide with the Lawn Hill Creek and Gregory subcatchments and Nicholson Groundwater Management Area (NGMA) .................... 67 Figure 4-2 Finite element mesh geometry showing pilot point locations and identifiers and specified head boundary conditions along Lawn Hill Creek and the Gregory River, and the throughflow boundary to the south ............................................................................................. 68 Figure 4-3 Location of groundwater head and flow reporting sites and the Nicholson Groundwater Management Area used for water balance reporting ........................................... 70 Figure 5-1 Depth to standing water level within the extended Southern Gulf catchments ........ 74 Figure 5-2 Spatial distribution of total dissolved solids (TDS) concentrations for major aquifers within the extended Southern Gulf catchments .......................................................................... 77 Figure 5-3 Cluster membership of aquifers in the Southern Gulf catchments. Total dissolved solids (TDS) corresponds to the median total dissolved solid concentrations of each cluster. The numbers correspond to the number of hydrochemical records assigned to each aquifer for major stratigraphic units ............................................................................................................... 79 Figure 5-4 Median ion values of the hierarchical cluster analysis presented as (a) Piper plot and (b) Schoeller plot ........................................................................................................................... 80 Figure 5-5 Spatial distribution of cluster membership within the extended Southern Gulf area 81 Figure 5-6 Spatial distribution of bore yields in major aquifers within the extended Southern Gulf catchments ............................................................................................................................ 84 Figure 5-7 The chloride (Cl) in groundwater (GW) observations within the study area and the median of the point-scale estimates of recharge derived from them ......................................... 85 Figure 5-8 Point-scale relationships between recharge and (a) rainfall, (b) clay content of the soil and (c) Normalised Difference Vegetation Index (NDVI) ....................................................... 86 Figure 5-9 Point-scale relationships between rainfall and recharge by geology class ................. 87 Figure 5-10 Coefficients used in the regression equations for upscaling the 1000 replicates (a–f), (g) the R2 for each of the 1000 replicates ..................................................................................... 88 Figure 5-11 (a) The median value of the 1000 replicates of upscaled recharge using the regression equation and (b) the median of the residuals kriged to a regular grid also showing the points used.............................................................................................................................. 89 Figure 5-12 The 5th, 50th and 95th percentiles of upscaled recharge from the 1000 replicates using regression kriging ................................................................................................................ 90 Figure 5-13 (a) Scatterplot of the 50th percentile of the recharge calculated at a point scale versus the 50th percentile of the regression-kriging upscaled recharge estimate (red line is 1:1 for reference), and (b) the R2 of all 1000 replicates for the point estimates of recharge versus the upscaled estimates of recharge as a boxplot ......................................................................... 90 Figure 5-14 The 5th, 50th and 95th percentiles of constrained recharge for the modelled area 91 Figure 5-15 The 5th, 50th and 95th percentiles of constrained recharge for the Southern Gulf catchments .................................................................................................................................... 92 Figure 5-16 Water bodies in the Southern Gulf catchments identified from Digital Earth Australia and the proportion of time that water bodies are inundated from Water Observations from Space .................................................................................................................................... 95 Figure 5-17 Excess water across the Southern Gulf catchments .................................................. 97 Figure 5-18 Areas of potential groundwater discharge across the Southern Gulf catchments ... 99 Figure 6-1 Thickness (isopach) map of Georgina Basin .............................................................. 102 Figure 6-2 Three-dimensional geological model (30× vertically exaggerated) of Georgina Basin showing (a) basement topography, groundwater divides and inferred flow direction; (b) Georgina Basin strata, springs and watercourses within Southern Gulf catchments; and (c) thickness of Georgina Basin ........................................................................................................ 103 Figure 6-3 Cross-section through the Georgina Basin. For orientation of cross-section, see Figure 6-1 ............................................................................................................................................... 104 Figure 6-4 Mineralogical composition (X-ray diffraction, XRD) of Georgina Basin and underlying strata at stratigraphic well Carrara 1 (based on data from Owen et al., 2023).......................... 105 Figure 6-5 Allaru Mudstone cores at Dobbyn 1 exploration well at a depth from approximately 232 to 245 m below ground surface ........................................................................................... 106 Figure 6-6 Gilbert River Formation core sample at Dobbyn 1 exploration well at a depth of approximately 581 m below ground surface. The inset is a zoom showing coarse quartz grains and open pore space in between quartz grains. ........................................................................ 106 Figure 6-7 Three-dimensional geological model of the Southern Gulf catchments showing (a) thicknesses and subsurface geometry of Gilbert River and overlying stratigraphic units; (b) spatial distribution of chloride mass balance recharge estimations; and (c) illustration of alternative subsurface geometry of Gilbert River Formation at the basin margin where the Gilbert River Formation is in direct contact with Cenozoic strata .............................................. 108 Figure 6-8 Thickness (isopach) map of the Gilbert River Formation .......................................... 109 Figure 6-9 Depth to the top of the Gilbert River Formation ....................................................... 110 Figure 6-10 Cross-section through the Carpentaria Sub-basin of the Great Artesian Basin within the Southern Gulf catchments. See Figure 6-9 for cross-section location ................................. 111 Figure 6-11 Mineralogical composition (X-ray diffraction, XRD) at Dobbyn 1 exploration well 111 Figure 6-12 Mineralogical composition (X-ray-fluorescence, XRF) at Dobbyn 1 exploration well ..................................................................................................................................................... 112 Figure 6-13 Block diagram of alluvium and sub-alluvial bedrock or older Cenozoic sediments along bore transect 1 near the confluence of the Gregory River and Nicholson River. The location of the transect is shown on Figure 6-7 ......................................................................... 113 Figure 6-14 Block diagram of the Gregory River alluvium and sub-alluvial older Cenozoic sediments (Karumba Basin) or bedrock along bore transect 2. The location of the transect is shown on Figure 6-7 .................................................................................................................... 113 Figure 6-15 Three-dimensional representation of regional groundwater heads and inferred groundwater flow lines in the Cambrian dolostone and limestone of the Georgina Basin ....... 114 Figure 6-16 Regional depth to water in the Georgina Basin ...................................................... 116 Figure 6-17 Three-dimensional geological model of the Georgina Basin and upscaled median recharge (Section 5.5) within the Southern Gulf catchments .................................................... 117 Figure 6-18 Minimum observed September streamflow at two stream gauge locations on the Gregory River .............................................................................................................................. 118 Figure 6-19 a) Topography (digital elevation model) of western part of the Southern Gulf catchments, streams and mapped springs and b) spatial extent of Constance Sandstone and Georgina Basin with inset cross-section along line A-B .............................................................. 119 Figure 6-20 Water-stable isotopes measured (δ2H vs δ18O) in groundwater located in Cambrian dolostone and limestone aquifer (adopted from Dixon-Jain et al., 2024) ................................. 121 Figure 6-21 Water-stable isotopes measured in groundwater. Low values are displayed in red while high values are displayed in blue ...................................................................................... 122 Figure 6-22 Carbon-13 vs carbon-14 in groundwater under the Undilla Sub-basin................... 123 Figure 6-23 Radiocarbon value in percent modern carbon (pMC) of the Cambrian dolostone and limestone aquifer displayed with the simplified geology as background .................................. 124 Figure 6-24 Great Artesian Basin (GAB) potentiometric surface and generalised flow lines..... 125 Figure 6-25 Selected groundwater-level hydrographs from the Gilbert River Formation ......... 126 Figure 6-26 Relationship between chloride (Cl⁻) concentrations and various major ions (Ca²⁺, Mg²⁺, Na⁺, HCO₃⁻, SO₄²⁻, Fe and SiO₂)......................................................................................... 127 Figure 6-27 Map displaying salinity by total dissolved solids (TDS) in milligram per litre values within the model survey area plotted with simplified geological units ..................................... 128 Figure 6-28 Sodium (Na) against chloride (Cl) in the Cambrian dolostone and limestone in the Undilla Sub-basin area ................................................................................................................ 129 Figure 6-29 Chloride (Cl) concentration in groundwater (milligrams per litre) in Undilla Sub- basin displayed by crop suitability to concentration sensitivity, plotted with simplified geological units ............................................................................................................................................. 130 Figure 6-30 Map of sulfate (SO4) concentration measure in Cambrian limestone and dolostone groundwater in Undilla Sub-basin (milligrams per litre) ............................................................ 131 Figure 6-31 Map of sodium adsorption ratio (SAR) concentration in millimole per litre displayed with the simplified geology in background ................................................................................. 132 Figure 6-32 Sodium adsorption ratio (SAR) and electrical conductivity (EC) calculated where colour represents total dissolved solids (TDS), the stability of soil is highlighted by two threshold lines displayed for soils composed by different clay concentrations (Line A: 55% to 65% and Line B: 25% to 35%) for an annual rainfall of 1000 mm/year ............................................................ 133 Figure 6-33 Piper diagram of Cambrian limestone and dolostone groundwater displaying only samples with an ionic balance less than 10% and a complete dataset (N = 213). The colour represents the associated total dissolved solids (TDS) ............................................................... 134 Figure 6-34 Chadha diagram in percent milliequivalents per litre plotted where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) ..................................................................................................................... 135 Figure 6-35 Ca and Mg versus bicarbonate concentration in milliequivalents per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) ......................................................................................................... 136 Figure 6-36 Ca versus SO4 concentration in milliequivalents per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) ...................................................................................................................................... 137 Figure 6-37 (Ca + Mg) – (Na + K) in percent milliequivalents per litre versus sulfate concentration in milligrams per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) ............................ 138 Figure 6-38 Saturation indices of minerals (A: gypsum, B: anhydrite, C: dolomite, D: aragonite, E: calcite) according to chloride (Cl) concentration in milligrams per litre, where colour represents the electrical conductivity for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) ...................................................................................................................................... 139 Figure 6-39 Mapping of dolomite saturation indices (SI) ........................................................... 140 Figure 6-40 Mapping of calcite saturation indices (SI) ............................................................... 141 Figure 6-41 Map of salinity in Gilbert River Formation and Rolling Downs Group .................... 143 Figure 6-42 Sodium (Na) against chloride (Cl) in the Gilbert River Formation and Rolling Downs Group with seawater dilution line in orange .............................................................................. 144 Figure 6-43 Map of depth to the top of the Gilbert River Formation and total dissolved solids (TDS) in the Gilbert River Formation ........................................................................................... 145 Figure 6-44 Map of the depth to the top of the Gilbert River Formation, bore depth to the top of the Rolling Downs Group and total dissolved solids (TDS) in the Rolling Downs Group ....... 146 Figure 6-45 Piper diagram of groundwater samples .................................................................. 147 Figure 6-46 (a) Ca, (b) Mg, (c) HCO3 and (d) SO4 against chloride (Cl) concentration in the Gilbert River Formation and Rolling Downs Group .................................................................... 149 Figure 6-47 (a) Calcite and (b) dolomite saturation indices against calcium (Ca) concentration in the Gilbert River Formation and Rolling Downs Group .............................................................. 150 Figure 6-48 (a) Na and (b) HCO3 against chloride (Cl) concentration in the Gilbert River Formation and Rolling Downs Group plotted showing the water type...................................... 151 Figure 6-49 Great Artesian Basin (GAB) potentiometric surface and generalised flow line ...... 153 Figure 6-50 Chloride (Cl), HCO3 and SO4 against distance along the generalised flow line using a 30 km buffer zone in Figure 6-49 ................................................................................................ 154 Figure 6-51 Sodium (Na), calcium (Ca) and magnesium (Mg) against distance along the generalised flow line using a 30 km buffer zone in Figure 6-49 ................................................. 155 Figure 7-1 Hydrogeological units with potential for future groundwater resource development ..................................................................................................................................................... 168 Figure 8-1 Example of isotope tracer sampling opportunities for the Gilbert River Aquifer in the Southern Gulf catchments. ......................................................................................................... 173 Tables Table 2-1 Lithostratigraphy of the Carpentaria Basin (Great Artesian Basin) sediments. Adapted from Orr et al. (2020) and Smerdon et al. (2012) ......................................................................... 26 Table 5-1 Overview statistics on bore construction and basic hydrogeological and hydrochemical characteristics of aquifers .................................................................................... 76 Table 5-2 Median values of the variables considered in the cluster analysis for each sample group ............................................................................................................................................. 78 Table 5-3 Mean recharge rates over each of the major aquifers. The 50th percentile is outside the brackets and the 5th and 95th percentiles give a range for the uncertainty within the brackets ......................................................................................................................................... 93 Table 5-4 Mean recharge rates over the simplified surface geology classes. The 50th percentile is outside the brackets and the 5th and 95th percentiles give a range for the uncertainty within the brackets .................................................................................................................................. 93 Table 5-5 Summary of areas identified as potential groundwater discharge areas..................... 98 Table 6-1 Mean, median, minimum and maximum water-stable isotope values for groundwater located in Cambrian dolostone and limestone aquifer in the extended Assessment area ....... 120 Table 6-2 Number of bores of the Na–HCO3, Na–Cl and Na–SO4 types in the Gilbert River Formation and Rolling Downs Group .......................................................................................... 148 Table 6-3 Geosciences Australia sample number, bore registration number, bore name, longitude, latitude and carbon isotope results for two Gilbert River Formation bores ............. 152 Table 6-4 Stratigraphic unit of the Rolling Downs Group bores ................................................. 156 Table 7-1 Opportunity-level estimates of the potential scale of groundwater resource development opportunities in the Southern Gulf catchments .................................................. 166 Part IIntroduction and overview Australian Fossil Mammal Site, Riversleigh section Photo: CSIRO 1 Introduction Section 1 of the report provide an introduction and overview, states the key aims of the study, provides a concise high-level summary of previous hydrogeological investigations and current water demand, and outlines the structure of the report. 1.1 Overview In northern Australia, future planning, investment and management of water resources relies on new or improved information across large areas, which can lack data. Future industries, communities and the environment depend on groundwater as a reliable year-round water source where surface water resources are either absent or ephemeral. Identifying opportunities and risks for future groundwater development requires understanding the coincidence of, and relationships between, suitable land, geology, productive and spatially extensive groundwater systems, existing water users and groundwater-dependent ecosystems (GDEs). In addition, evaluating potential hydrological impacts on existing groundwater users and ecologically and/or culturally important GDEs requires understanding the influences on hydrological processes of the distinctive seasonality and large inter-annual variability in rainfall and potential evaporation. In many places across the catchments of the Southern Gulf rivers, that is Settlement Creek, Gregory–Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups, rivers and creeks are either intermittent or ephemeral. Communities source their water equally from either surface water or groundwater for a variety of purposes. These include uses for: (i) stock and domestic purposes, (ii) town and community water supplies, and (iii) industries such as agriculture and mining. The major regional aquifers in the Southern Gulf catchments are hosted in the interconnected Cambrian dolostone and limestone (including the Camooweal Dolostone and Thorntonia Limestone) and in the Gilbert River Formation, the major Late Cretaceous to Early Jurassic aquifer of the Carpentaria Basin (a sub-basin of the Great Artesian Basin). Local aquifers likely also occur in the Cenozoic alluvium, Proterozoic igneous rocks, and Proterozoic metasedimentary and metamorphic rocks. As the nature of each aquifer varies across the catchments with spatial changes in geology and hydrogeology, it is important to assemble existing and collect new information to provide a more robust spatial understanding of the groundwater flow systems they host. Key characteristics include their: (i) extent and depth in the subsurface, (ii) thickness, (iii) storage and flow, and (iv) connectivity with other aquifers and aquitards, and with subterranean, aquatic and terrestrial GDEs. GDEs include groundwater-fed rivers and creeks, riparian vegetation, wetlands, waterholes, springs, spring and groundwater-fed vegetation, stygofauna and the marine environment. Consequently, identifying and evaluating promising aquifers with potential for future groundwater resource development and providing information to guide groundwater planning, investment and management involves: • identifying and characterising the depth and spatial extent of target aquifers • conceptualising the nature of their flow systems • estimating water balances • estimating potential extractable volumes from the aquifers and the associated drawdown in groundwater level or changes in flow over time and distance at various locations throughout each aquifer. This information will facilitate groundwater resource planning by describing current baseline conditions and how these may be subject to future changes to groundwater levels and flow in aquifers at particular locations. Planning decisions require value judgments as to what is an acceptable impact (in terms of change to the natural flow regime) on receptors such as environmental assets or existing users at a given location. These decisions can be complex and require considerable input from a wide range of stakeholders, particularly government regulators and communities. 1.2 Aims of the groundwater hydrology activity The purpose of the groundwater hydrology activity in the Southern Gulf Water Resource Assessment was to examine opportunities for future groundwater development to support primary industries in the Southern Gulf catchments, principally irrigated agriculture but potentially also aquaculture. At the scale of the Southern Gulf catchments, the activity identified and assessed the most promising intermediate to regional-scale aquifers and where sufficient information exists, quantified the potential opportunities for, and risks associated with, future groundwater development. The key questions that this activity seeks to address in the Southern Gulf catchments include: • What types of aquifers exist and what is the nature of the flow systems they host? • What are the important attributes that help identify aquifers in the catchments as promising for future groundwater resource development and how do they vary spatially? • Can a range in recharge for these aquifers be estimated and are these ranges reasonable when considering rainfall, runoff, evapotranspiration and groundwater levels? • Which river or creek reaches have evidence of strong groundwater–surface water connectivity and which aquifers support their persistence? • What types of vegetation are utilising groundwater for transpiration, where do they occur, and which aquifers support their water use? • For the most promising aquifers identified, what component of their water balance can be estimated with some level of confidence and what components remain uncertain? • What are the ranges in potential extractable volumes for the most promising aquifers and what are the key risks posed by future groundwater extraction? 1.3 Summary of previous hydrogeological investigations The Jurassic to Cretaceous-aged Carpentaria Basin is comprised of interbedded sandstones, mudstones and siltstones, which underlie most of the north-eastern parts of the Southern Gulf catchments. According to Smerdon et al. (2012), the Carpentaria Basin is separated from adjoining geological basins by basement highs, and together with a number of other geological basins located to the south of the Southern Gulf catchments, including the Eromanga and Surat basins and parts of the Clarence-Moreton Basin, collectively form parts of the Great Artesian Basin (GAB). Fluvial quartzose sandstones began to deposit in the Carpentaria Basin during the Jurassic, mostly restricted to palaeo-topographic valleys (Orr et al., 2020). Previous studies identified five main types of aquifers across the Southern Gulf catchments: (i) fractured and weathered rocks; (ii) fractured, fissured and karstic carbonate rocks of the Georgina Basin; (iii) extensive porous sedimentary sandstones of the GAB; (iv) porous sandstones of the Karumba Basin; and (v) surficial unconsolidated to consolidated alluvial sands and gravels (CSIRO, 2009a; 2009b; 2009c; Smerdon et al., 2012). Most bores intersecting the fractured, fissured and karstic carbonate rocks of the Georgina Basin range in depth from 50 to 100 mBGL, though standing water levels (SWLs) can be deep (i.e. between 70 and 100 mBGL) ((Department of Environment, Parks and Water Security, 2021). Indicative bore yields for the karstic aquifers derived from airlifting or submersible pumps at the time of drilling appear reasonable, ranging up to 14 L/second with most bores yielding 0.5 to 5 L/second (Matthews, 1992; Read, 2003). Data for the aquifers’ hydraulic properties is limited and is expected to be highly variable given the karstic nature of the aquifers. Investigations by Matthews (1992) and data collated by Read (2003) suggest that transmissivities of the dolostone range between 290 and 1700 m2/day. Localised spring discharge and diffuse seepage from the Camooweal Dolostone supports dry- season baseflow to the Gregory River and its tributaries as well as Lawn Hill Creek in the Southern Gulf catchments (CSIRO, 2009c; Jolly and Tickell, 2011; Tickell, 2003), with dry-season flow rates in the Gregory River and its tributaries estimated to be between 150 and 350 million L/day (Jolly and Tickell, 2011). Both of these groundwater-dependent streams are listed as nationally significant wetlands (Environment Australia, 2001) and they are recognised as areas of high cultural significance, with Indigenous occupation at Lawn Hill dating back at least 17,000 years and possibly extending beyond 30,000 years (Jolly and Tickell, 2011). Jolly and Tickell (2011) also described the tufa deposits in the headwaters of the Gregory River and Lawn Hill Creek of the Southern Gulf catchments. They suggested that springs discharge from the Cambrian Thorntonia Limestone and Camooweal Dolostone where the Gregory River and its tributaries have incised into these formations. They explained that tufa deposits gradually built up to form small dams where rivers flow over obstructions such as rocks or vegetation debris. As part of the Australian Government Geological and Bioregional Assessment (GBA), Orr et al. (2020) and Buchanan et al. (2020) assessed the hydrogeology and groundwater systems of the Isa GBA region within the Southern Gulf catchments. This desktop study compiled geological and hydrogeological information and provided a comprehensive overview on groundwater and surface water systems within the Isa GBA region, located between Burketown and Doomadgee, which corresponds to approximately 20% of the Southern Gulf catchments. An inventory of hydrochemical and environmental tracer data conducted as part of Geoscience Australia’s project, ‘Assessing the Status of Groundwater in the Great Artesian Basin’, was created across aquifers of the GAB by Raiber et al. (2022). The compilation highlighted the lack of environmental tracer data within the Carpentaria Basin (a sub-basin of the GAB), where only a small number of environmental tracer records exist, and it was therefore deemed as a key knowledge gap. Crosbie et al. (2022) used the chloride mass balance (CMB) method to estimate recharge across the entire GAB. This study identified the Carpentaria Basin, including the northern part of the Carpentaria Basin, as an area of high recharge. However, it also highlighted the lack of data underpinning the assessment in this area. This recharge study also informed the GAB-wide study on ‘Assessing the Status of Groundwater in the Great Artesian Basin’. In the summary report of this project, Wallace and Ransley (2022) presented the results of a water balance for key GAB aquifers. They suggested that the northern part of the Carpentaria Basin experiences high recharge and relatively low groundwater use, whereas the annual rainfall is lower, and groundwater is an increasingly important water source in the southern part (which includes part of the Southern Gulf catchments). The authors suggested that some hydrographs show declining trends, reflecting an overall water balance result of declining storage for the Carpentaria Basin. As part of the same project, Vizy and Rollet (2023a, 2023b) developed updated three-dimensional geological model surfaces of the major aquifers of the GAB. A recently published report by Geoscience Australia (Dixon-Jain et al., 2024) assessed the regional geology, hydrogeology and groundwater systems of the South Nicholson–Georgina region in the NT and Queensland. This study presented a whole-of-basin conceptualisation of groundwater flow systems and recharge and discharge processes within the regional unconfined aquifers of the Georgina Basin. It differentiated the Georgina Basin into three groundwater flow systems separated by basement highs. The study also conceptualised the connectivity between aquifers and springs in the Lawn Hill area of the Southern Gulf catchments using airborne electromagnetic (AEM) data. As part of this study, Vizy and Rollet (2024) developed new three-dimensional geological model surfaces of the Georgina Basin. In summary, these previous hydrogeological investigations indicate that the regional-scale porous aquifers of the Carpentaria Basin and karstic carbonate aquifers of the Georgina Basin may offer potential for future groundwater resource development though are data sparse. 1.4 Current water demand and licensed entitlements Communities and industries in the Southern Gulf catchments source their water equally from either surface water or groundwater for a variety of purposes. These include uses for: (i) stock and domestic purposes, (ii) town and community water supplies, and (iii) industries such as agriculture and mining. Where surface water is used in some applications, water is pumped from the occasional dam or stream. In the case of Mount Isa’s water supply, a major water transmission pipeline supported by pumping stations is used to transfer water from the dams on Lake Julius and Lake Moondarra for treatment prior to distribution in Mount Isa. Where groundwater is used in some applications, small quantities of groundwater (i.e. <100 ML/year) may be pumped from a single bore for stock and domestic use. In cases where larger amounts of groundwater are required (i.e. >100 ML/year) water may be pumped from a bore field consisting of multiple connected production bores in applications such as town and community water supplies or irrigated agriculture. Some applications of water use are associated with a water licence such as water use for town and community water supply, or applications by industry. Other applications such as stock and domestic use do not require a licence. Water use in the catchments occur in some water planning areas. For surface water this may include the Gulf Water Plan area. For groundwater this may include the Great Artesian Basin and Other Regional Aquifers Plan area. The Great Artesian Basin and Other Regional Aquifers Water Plan manages groundwater sources from multiple aquifers hosted in different geological units and within different groundwater sub-areas (Figure 1-1). 1.4.1 Surface water entitlements Surface water licences with a volumetric entitlement occur at a variety of locations and from a variety of sources across the Southern Gulf catchments (Figure 1-1). Twenty-seven surface water licences with a volumetric entitlement have been granted for a combination of use for agriculture and aquaculture, across various parts of the catchments (Figure 1-1) (Department of Regional Development Manufacturing and Water, 2021). The largest entitlements (i.e. between 1000 and 8000 ML/year) have been granted for agricultural use. Some moderate entitlements (i.e. between 400 and 1000 ML/year) have been granted for town and community water supply in Mount Isa, Gregory and Kajabbi (Figure 1-1). Much smaller surface water entitlements (i.e. <50 ML/year) are associated with stock use (Figure 1-1). 1.4.2 Groundwater entitlements There are currently 13 groundwater licences with a volumetric entitlement that have been granted for a variety of applications (Figure 1-1) (Department of Regional Development Manufacturing and Water, 2021). The largest entitlements (i.e. 150 to 1400 ML/year) are associated with industrial use in mining, with the water sourced from various aquifers hosted in the Paradise Creek Formation and the Currant Bush and Thorntonia limestones (Figure 1-1). Two licensed entitlements of approximately 100 ML/year have been granted for town and community water supplies at Burketown and Gununa (Mornington Island). Both licences have been granted for groundwater sources from the Gilbert River Formation of the GAB (Figure 1-1). The smallest groundwater licences (i.e. <100 ML/year) have been granted for a variety of industrial and agricultural uses, with groundwater sourced from a variety of aquifers hosted in different geological units. The Century Zinc Mine, located about 15 km to the south-east of Lawn Hill (Figure 1-1), was Australia’s largest open-pit zinc mine before its closure in 2016. It used to de-water part of the Cambrian Limestone Aquifer (CLA) hosted in the Thorntonia Limestone that overlies zinc deposits hosted in the Proterozoic Lawn Hill Formation. When fully operational, the mine was reported to be extracting about 19 GL/year of groundwater in the early 2000s and about 10 GL/year in the mid-2010s. Currently, less than 1 GL/year of groundwater is being extracted at the site. The cessation in de-watering at the site is likely to have resulted in recovery of groundwater levels and storage in the CLA around the site. This may also include an onset of increased discharge from the aquifer to Lawn Hill Creek. However, the timescales for changes in groundwater flow are likely to be in the order of tens of years or longer, and further investigation would be required to confirm this. Figure 1-1 Spatial distribution of surface water and groundwater licensed entitlements across the Southern Gulf catchments Water license data source: Department of Regional Development Manufacturing and Water (2021). 1.5 Report overview and structure This report describes the methods and tools used to conduct a regional assessment of hydrogeological systems of the Southern Gulf catchments and more detailed desktop and modelling investigations of the Cambrian Limestone Aquifer (CLA) hosted in the Cambrian dolostone and limestone of the Georgina Basin and the Gilbert River Aquifer (GRA) hosted in the Cretaceous to Early Jurassic Gilbert River Formation of the Carpentaria Basin. The report is structured as follows: • Section 2 outlines the geography, demography, climate, geology and hydrogeology of the study area. • Sections 3 and 4 detail the methods and tools used to conduct the regional hydrogeological assessment (water levels, bore yields, water quality, recharge estimation and discharge mapping) and targeted field, desktop and modelling investigations, respectively. • Sections 5 and 6 present and discuss the results of the regional assessment and desktop and modelling investigations, respectively. • Section 7 discusses the opportunities and constraints associated with further development of groundwater resources from the CLA and GRA and from other aquifers in the Southern Gulf catchments. • Section 8 states the conclusions of this study and potential options for future work. 2 Study area Section 2 of the report describes the study area including the: (i) physiography and demography, (ii) climate, (iii) geology and hydrogeology, (iv) surface water hydrology, and (v) water dependent ecosystems. 2.1 Physiography and demography The catchments of the Southern Gulf rivers, that include Settlement Creek, Gregory–Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups, span an area of 108,200 km2 across the NT and Queensland. They are comprised of the Settlement (17,600 km2), the Nicholson (52,200 km2), the Leichhardt (33,400 km2), and the Morning Inlet (3,700 km2) catchments, and the Mornington Islands (1,200 km2), shown by the black outlines in Figure 2-1. The catchments exhibit ten different physiographic units, which are described in detail in Grimes (1974) and have been adapted to be displayed in Figure 2-1. The catchments can generally be described as the uplands (highlands and plateaux) and the coastal plains (Grimes, 1974). The upland area in the south and west of the catchments reaches a maximum of about 620 mMSL (metres above mean sea level) and hosts the headwaters for the major streams in the catchments. The uplands can be divided into four physiographic units shown in Figure 2-1. The oldest, most elevated and rugged unit is the Isa Highland (Twidale, 1956). It consists of Precambrian volcanic and sedimentary rocks that have been metamorphosed, weathered and eroded. Land surface elevation in this part of the catchments is moderate (200 to 230 mMSL) and generally has a south to north alignment. The next most elevated uplands occur on a small part of the Barkly Tableland west of the Isa Highland. The tableland started out as a sedimentary basin in Precambrian times, which was subsequently uplifted, folded and eroded. During the early Cambrian, seas transgressed the area and deposited carbonate sediments in the depressions. Later into the Cambrian, sediments were exposed and eroded. During the Mesozoic, isolated lakes and swamps developed (Randal, 1966), subsequently during the early Cenozoic the upland areas experienced deep weathering and laterisation. Since the Cambrian period, the drainage network flowing towards the Gulf of Carpentaria has dissected the tableland leaving remnant land features defined by deep narrow gorges. This area is mapped as the Dissected Barkly Tableland in Figure 2-1. Dissection has been amplified because the underlying rocks formed from dolomitic sediments are relatively soluble compared to surrounding rocks. These gorges have intersected the groundwater systems of the tableland resulting in spring- fed perennial streams such as the O’Shannassy and Gregory rivers, and Lawn Hill Creek. The remaining parts of the uplands are comprised mainly of Mesozoic sandstones, which have been eroded into a complex pattern of easterly flowing streams and valleys separated by ranges and rocky outcrops (Smith and Roberts, 1972). The Nicholson and South Nicholson rivers are the primary systems draining this area known as the Gulf Fall (Figure 2-1). Musselbrook, Lagoon, Settlement, Gold and Running creeks also drain this part of the catchments. Figure 2-1 Physiographic regions of the Southern Gulf catchments Physiographic regions: Adapted from Grimes (1974) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au To the east of the uplands are the Carpentaria Plains comprising a series of plains, pediments (shallow slope at the foot of a steeper slope) and remanent plateaux that can be divided into six physiographic units. The most elevated sedimentary plain (30 to 150 mMSL) immediately east of the uplands is the Cloncurry Plain (Figure 2-1). It consists of gently sloping colluvial and fluvial sedimentary plains and pediments with isolated low hills of Precambrian rocks. Streams are sparse and incised into the pediments with narrow alluvial plains (Grimes, 1974). The Cloncurry Plain unit extends from the middle reach of the Leichhardt River to Lawn Hill Creek. In the north, the Doomadgee Plain lies below and adjacent to the Cloncurry Plain and is predominantly a sandy, gently undulating plain overlying a deeply weathered Cenozoic land surface (Figure 2-1). Widely spaced creeks drain the plains currently in a radial north-westerly direction towards the coast. In the south, the Armraynald Plain unit lies below and adjacent to the Cloncurry Plain and consists of argillaceous Cenozoic sediments (Armraynald Beds) (Figure 2-1). Stream channels are sparse, widely spaced and deeply incised due to sea-level changes. The plains extend up the Lawn Hill Creek, and the Gregory and Leichhardt valleys. Lawn Hill Creek and Gregory River are fed by perennial springs that also support spring-fed vegetation. The Gregory River is highly braided in this area with permanent streams consisting of the Gregory River, Beames Brook, Barkly River and Running Creek downstream of the Gregory Crossing (Grimes, 1974). Down the topographic gradient of the Doomadgee and Armraynald plains lies the coastal Karumba Plain (Figure 2-1). This coastal unit extends 10 to 35 km inland from the Gulf of Carpentaria coast, and the plain is most extensive near the mouth of the Albert River (Nicholson/Leichhardt River). This plain consists of Holocene beach ridges and tidal and extratidal flats and plains. These inland plains only flood occasionally. Because the plain is wide and generally flat, and the tidal range is moderate (about 3.5 m), tidal waters can rapidly inundate the land. Mangroves and tidal flats dominate the coastline, beaches are sparse and are comprised mostly of white shelly sand. Small crescent dunes have formed in places from wind action (Grimes, 1974). To the east of the Armraynald Plain unit is the Donors Plateau (Figure 2-1). This slightly elevated unit (10 to 80 mMSL) forms a watershed between the catchments of the Leichhardt and Flinders rivers and forms the eastern boundary of the Southern Gulf catchments. The plateau consists of siliceous sediments deposited during the Early Cretaceous from upland sediment sources of the Normanton Formation. The plain, which was once more extensive, has been deeply weathered and lateralised, and has subsequently been stripped away in parts. Much of the Wellesley Islands in the Gulf of Carpentaria represent remnants of a mainland laterised Cretaceous plain called the Mornington Plateau (Figure 2-1). This plateau is only about 5 to 20 mMSL. Dissection of the plain is not as extensive compared with the Donors Plateau. The Mornington Plateau is generally fringed with marine plains consisting of coastal sediments or dune fields lower in the landscape that support small mangroves, while more elevated areas host sea cliffs and wavecut platforms (Grimes, 1974). Agricultural production is the largest land use in the catchments, mostly cattle grazing on native pastures (85%) (Department of Environment, Parks and Water Security, NT Government, 2022; Queensland Government, 2021). Other land uses include recreational activities, tourism, traditional and commercial fisheries, mining and Indigenous uses. In addition, these catchments have important ecological and environmental values. Within these catchments and the surrounding marine environment, they host numerous important ecological assets (see Section 2.6). The population density of the Southern Gulf catchments is sparse at one person per 4.8 km2, which is about one-fourteenth that of Queensland and one-sixteenth that of Australia as a whole. The catchments contain one significant urban area (population >10,000 people), namely, Mount Isa, a city of over 18,000 residents that was developed to support mineral exploration and extraction (particularly lead, silver, copper and zinc). There are also numerous small towns and communities within the catchments, including Burketown and Doomadgee, and the Wellesley Islands. Of these smaller settlements, only Doomadgee has a population greater than 1000 people (Australian Bureau of Statistics, 2021). 2.2 Climate Weather, which is defined as short-term atmospheric conditions, is the key source of uncertainty affecting hydrology and crop yield. It influences the rate and vigour of crop growth with catastrophic weather events able to cause extensive crop losses. Climate is defined as weather of a specific region averaged over a long period of time. Key climate parameters controlling plant growth and crop productivity include rainfall, temperature, radiation, humidity, and wind speed and direction. Of all the climate parameters affecting hydrology and agriculture in water-limited environments, rainfall is usually the most important. Rainfall is the main determinant of runoff and recharge and is a fundamental requirement for plant growth. For this reason, reporting of climate parameters is heavily biased towards rainfall data. Other climate variables affecting crop yield are discussed in the companion technical report on climate (McJannet et al., 2023). Climate data presented in this report were calculated using SILO (Scientific Information for Land Owners) climate data surfaces (Jeffrey et al., 2001) unless stated otherwise. Very few climate data are available in the study area before 1890, therefore the 132-year period between 1 September 1890 to 31 August 2022 is used in the analysis presented below. Unless otherwise stated, the material in Section 2.2 is based on findings described in the companion technical report on climate (McJannet et al., 2023). 2.2.1 Weather patterns over the Southern Gulf catchments The Southern Gulf catchments are characterised by distinctive wet and dry seasons due to their location in the Australian summer monsoon belt (Figure 2-2). During the build-up months (typically September to December), the Southern Gulf catchments typically experience low-level easterly winds, which can carry pockets of dry or humid air and result in short-lived thunderstorm activity under favourable conditions. Over inland areas of the Southern Gulf catchments, storms form more frequently during the afternoon because of increased air temperature, which enhances instability and leads to convective cloud formation. Storms also form more readily near the heat trough that is a semi- permanent feature over inland Queensland (predominantly located south of the catchments) during late spring and the summer months. There is also a high incidence of thunderstorms in the Southern Gulf catchments where sea breeze convergence and/or boundaries act as a trigger. Thunderstorms show a strong diurnal variation, with most occurring during the afternoon and early evening. Dynamic forcing can cause thunderstorms to develop or persist well beyond the normal diurnal cycle, and if the dynamics are strong enough, thunderstorms can occur at any time. Figure 2-2 Historical rainfall, potential evaporation and rainfall deficit Median (a) annual, (b) wet-season and (c) dry-season rainfall; median (d) annual, (e) wet-season and (f) dry-season potential evaporation; and median (g) annual, (h) wet-season and (i) dry-season rainfall deficit in the Southern Gulf catchments. Rainfall deficit is rainfall minus potential evaporation. Data source: McJannet et al. (2023) "\\FS1-CBR.nexus.csiro.au\{lw-rowra}\work\1_Climate\4_S_Gulf\1_GIS\1_Map_docs\1_Exports\Cl-SG-507-Hist-MedAnnRF-ET-RFdeficit.png" For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au In the Southern Gulf catchments, convection and rain from sea breeze convergence typically occur just inland from the coast from early afternoon in the warmer months. Rain can continue overnight and into the following morning if conditions are favourable, particularly if there is a moisture feed from a strong easterly nocturnal jet stream. During the day, the air over land warms and rises, resulting in lower pressure at the surface. Air flows from the Gulf of Carpentaria waters towards the land to fill this area of lower pressure, resulting in the sea breeze, which can converge with the usually south-easterly synoptic winds. During the wet season, the convergence of persistent easterly winds and broader synoptic north-westerly winds over the Gulf of Carpentaria results in convergence, which may trigger storms over the coast after sunset. This phenomenon is most pronounced in the near-coastal areas of the catchments. The mean annual rainfall, averaged over the Southern Gulf catchments for the 132-year historical period, is 602 mm. Rainfall totals are highest near the coast and decline in a southerly direction (Figure 2-2). This is because the more northerly regions of the catchments receive more wet- season rainfall due to active monsoon episodes. The Southern Gulf catchments are relatively flat, so there is no noticeable topographic influence on climate parameters such as rainfall or temperature. Approximately 94% of annual rain in the Southern Gulf catchments falls during the wet-season months (1 November to 30 April). Figure 2-2 shows the spatial distribution of rainfall over the year and during the wet and dry seasons. Median wet-season rainfall exhibits a very similar spatial pattern to median annual rainfall. Median dry-season rainfall is highest in the most southern part of the Leichhardt catchment and is lowest near the coast. The highest monthly rainfall totals typically occur during January and February (Figure 2-2). 2.3 Geology 2.3.1 Major geological basins and provinces This section presents a summary of the depositional and tectonic history of geological basins in the Southern Gulf catchments, providing context relevant to this groundwater hydrology activity. The geological past of the Southern Gulf catchments is intricate and dominated by six major sedimentary basins overlying Proterozoic basement (Figure 2-3). In order of depositional history, these are the McArthur, the Isa Super, the South Nicholson, the Georgina, the Carpentaria and the Karumba basins (CSIRO, 2009c; Orr et al., 2020). Figure 2-4 also shows the entire spatial extent of the Georgina Basin and Carpentaria Sub-basin of the Great Artesian Basin (GAB) relative to the Southern Gulf catchments. Figure 2-3 Major geological provinces of the Southern Gulf catchments Source: adapted from Raymond et al. (2018) A map of the continent Description automatically generated Figure 2-4 Full extent of the Georgina Basin and Carpentaria Sub-basin of the Great Artesian Basin. Inset map shows full extent of Great Artesian Basin Geology data sources adapted from: Ransley et al. (2015); Carr et al. (2016) A map of the united states Description automatically generated Supported by new seismic surveys and drilling of stratigraphic wells conducted as part of the National Drilling Initiative (Australian Government website ) and by Geoscience Australia’s Exploring for the Future program (Carr et al., 2019), the understanding of the relationships between the McArthur Basin, the Mount Isa Province (which includes the Isa Superbasin) and the South Nicholson Basin continues to evolve. Geometric relationships of adjacent sedimentary basins in the Southern Gulf catchments in Queensland are shown in Figure 2-5. Figure 2-5 Geological cross-section (west to east) through the Southern Gulf catchments (Buchanan et al., 2020) GBA = Geological and Bioregional Assessment (from the Australian Government Geological and Bioregional Assessment Program). The McArthur Basin was deposited during the late Palaeoproterozoic to early Mesoproterozoic (approximately 1780 to 1400 Ma). It hosts some of the oldest rocks within the Southern Gulf catchments. However, it only underlies a very small portion of the north-west of the catchments (mostly beneath Settlement Creek catchment) and is comprised of a succession of sandstone, shale, carbonate, and interbedded volcanic and intrusive igneous rocks (Withnall and Cranfield, 2013). Most of the oldest rocks in the Southern Gulf catchments are associated with the Isa Superbasin and the South Nicholson Basin. The former is one of four superbasins identified in the North Australian Craton that formed during the Palaeoproterozoic–Mesoproterozoic (1670 to 1575 Ma; Bradshaw et al., 2018; Orr et al., 2020). Superbasins are described as originally very large single depositional systems disrupted internally by later-stage tectonic deformation resulting in For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au numerous structural basins and other structural features (Buchanan et al., 2020). The Isa Superbasin hosts sequences of dolostone and limestones, and siliciclastic and volcanic rocks of the lower and upper McNamara Group (Buchanan et al., 2020). The Isa Superbasin has been primarily defined and described within the Lawn Hill Platform, a relatively undeformed part of the Mount Isa Inlier, which has been subjected to low, greenschist grade metamorphism (Buchanan et al., 2020). The principal stratigraphic group of the Mesoproterozoic South Nicholson Basin (1483 to 1266 Ma; Anderson et al., 2019) is the South Nicholson Group, which may be an age equivalent of the Roper Group of the NT (Bradshaw et al., 2018). The South Nicholson Basin sequence is composed entirely of sedimentary rocks that were deposited in fluvial to shallow marine environments (Orr et al., 2020). The South Nicholson region extends through parts of the Southern Gulf catchments in both Queensland and the NT. It has been described as the subsurface extent of the sedimentary successions composed of the Isa Superbasin, the overlying South Nicholson Group and their lateral equivalents (Bailey et al., 2020). Within the eastern part of the Southern Gulf catchments in Queensland, the sedimentary successions of the Isa Superbasin and the South Nicholson Group occur at great depths of typically more than 1 km and up to approximately 9 km, but they subcrop or outcrop in the western part of the Southern Gulf catchments in Queensland (Figure 2-5; Buchanan et al., 2020; Orr et al., 2020). The Neoproterozoic to late Palaeozoic (850 to 350 Ma) Georgina Basin occurs in the south-western portion of the Southern Gulf catchments. It has been described as a relatively thin stratigraphic sequence of mostly shale, limestone, dolomite and minor sandstone (Smith, 1972). These sediments were deposited on a tectonically inactive platform, with a marine transgression from early Middle Cambrian to Late Cambrian and are largely concealed by overlying Cretaceous sediments. Drilled to a depth of 1751 m as part of the South Nicholson National Drilling Initiative (Australian Government website ) in the south-west of the Southern Gulf catchments (Figure 2-6), the new stratigraphic well Carrara 1 provides insights into the geology of the Georgina and South Nicholson basins. It demonstrates that at this location, the Georgina Basin hosts a sequence of approximately 580 m of carbonate rocks of the Camooweal Dolostone and the Currant Bush Limestone and minor dominantly siliciclastic rocks of the Border Waterhole Formation (Figure 2-6). Beneath the Georgina Basin, approximately 1100 m of Proterozoic carbonates, black shales and minor siliciclastic of the Lawn Hill Formation and Plain Creek Formation were found. Figure 2-6 Composite logs of Carrara 1 stratigraphic well (Geoscience Australia, 2023) Within the Southern Gulf catchments in Queensland, a substantial gap in the sedimentary record exists after deposition of the Mesoproterozoic South Nicholson Group. The Jurassic to Cretaceous-aged Carpentaria Basin is comprised of interbedded sandstones, mudstones and siltstones, which underlie Cenozoic sediments across most of the north-eastern parts of the Southern Gulf catchments. According to Smerdon et al. (2012), the Carpentaria Basin is separated from adjoining geological basins by basement highs, and together with a number of other geological basins located to the south of the Southern Gulf catchments, including the Eromanga and Surat basins and parts of the Clarence-Moreton Basin, collectively form parts of the Great Artesian Basin (GAB) (Smerdon et al.,2012); Wallace and Ransley, 2022). Fluvial quartzose sandstones began to deposit in the Carpentaria Basin during the Jurassic, mostly restricted to palaeo-topographic valleys (Orr et al., 2020). This includes deposits of the Gilbert River Formation, which is also comprised of fluvial quartzose sandstones, followed by a widespread transgression and then major regression in the late Middle Cretaceous, resulting in the thick mudstone successions of the upper Wallumbilla Formation (CSIRO, 2009a). Open marine conditions began to dominate to the north during the Cretaceous period associated with the final Cretaceous regression of the Eromanga Sea northwards (Foley et al., 2022), resulting in gradually decreasing organic content in the overlying Toolebuc Formation. In the southern Carpentaria Basin, the fine- grained siltstones and mudstones of the Allaru Mudstone were likely deposited under shallow marine conditions (Buchanan et al., 2020). The Normanton Formation, which caps the GAB sequences in the Carpentaria Basin, was deposited in a large, northward-protruding, likely river- dominated delta system and is composed of volcanoclastic material (Foley et al., 2022). The Normanton, Toolebuc and Wallumbilla formations and the Allaru Mudstone are collectedly known as the Rolling Downs Group (Smerdon et al., 2012). Extensive Cenozoic-aged alluvial plains deposits unconformably overlie the Carpentaria Basin and mostly correspond to the Cenozoic Karumba Basin (Orr et al., 2020), which is persistent across much of northern Queensland and extends into eastern parts of the NT (Cook and Jell, 2013). Three major cycles of erosion, deposition and weathering affected this shallow intracratonic basin For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au preceded by active tectonics, which provided erosional surfaces that were transported. The Bulimba Formation was associated with the first cycle and corresponds to interbedded conglomerates, sandy claystone and clayey sandstones covered by a deep siliceous duricrust formed during the weathering phase. Overlying the Karumba Basin are the youngest sediments in the catchments, the alluvial sands, silts and gravels associated with the beds, channels and floodplains of the catchments’ rivers and creeks, and their tributaries. Only few groundwater bores intersect the youngest alluvial aquifers. These are generally relatively shallow (drilled to depths of up to approximately 25 m) (Buchanan et al., 2020). 2.3.2 Surface geology Geological history represents the major periods of deposition and tectonics (i.e. major structural changes), as well as weathering and erosion. These processes are closely linked to the physical environment that influences the evolution and formation of resources such as valuable minerals, coal, groundwater and soil. Geology also controls topography, which in turn is a key factor that influences hydrological processes such as runoff, streamflow, flooding and recharge to groundwater. The oldest rocks in the Southern Gulf catchments are late Proterozoic (1780 to 1400 Ma) and consist of repeated thick sequences of sedimentary and metamorphic rocks and volcanics that include numerous prominent beds of sandstone (Figure 2-7). They were deposited in a series of basins (e.g. the McArthur, Isa and South Nicholson basins) extending across the area and then folded, faulted and intruded by igneous rocks to form mountain chains. Towards the end of the Proterozoic, the mountain chains had been eroded down to a level not far above that of the current topography. During the Neoproterozoic to late Palaeozoic (850 to 350 Ma), the limestones, dolomites and minor sandstones of the Georgina Basin were deposited on a tectonically inactive platform. The Cambrian strata provide an important regional groundwater source in the west and south-west of the Southern Gulf catchments and are partly concealed by overlying Cretaceous to Cenozoic sediments (Figure 2-7). The Jurassic to Cretaceous-aged Carpentaria Basin (125 to 100 Ma), a sub-basin of the GAB, hosts sequences of interbedded sandstones, mudstones and siltstones, which underlie most of the eastern part of the Southern Gulf catchments and extend and thicken offshore. This includes deposits of the Gilbert River Formation, which is comprised of fluvial quartzose sandstones and forms the major GAB aquifer and an important groundwater resource within the north-eastern parts of the Southern Gulf catchments (Figure 2-7). Following the deposition of the Gilbert River Formation, widespread transgression and then major regression in the late Middle Cretaceous led to deposition of the thick mudstone successions of the Wallumbilla Formation and erosion and deposition of a thin succession of Cretaceous shallow marine sandstone, conglomerate and mudstones. Extensive Cenozoic-aged alluvial plains deposits unconformably overlie the Carpentaria Basin in the north of the catchments and mostly correspond to the Cenozoic Karumba Basin, which is persistent across much of northern Queensland and extends into eastern parts of the NT. Figure 2-7 Surface geology of the Southern Gulf catchments Surface geology data source: Raymond et al., 2012 A map of the country Description automatically generated Overlying the Karumba Basin are the youngest sediments in the catchments, the alluvial sands, silts and gravels associated with the beds, channels and floodplains of the catchments’ rivers and creeks, and their tributaries. The present landscape has been produced by warping and dissection of a series of erosion surfaces formed during several cycles of erosion that started in the Late Cretaceous about 70 Ma and ended in the mid-Cenozoic era about 25 Ma. During this time, stable crustal conditions and subaerial exposure led to patchy erosion of the Cretaceous rocks and prolonged subaerial weathering of the remaining Cretaceous and Proterozoic rocks, resulting in the formation of deep weathering profiles and associated iron-cemented capping. Between the mid-Cenozoic and the present day, there has been gentle uplift and warping of the various surfaces and their weathered capping. Continued erosion has led to the emergence of the present-day landscape, and extensive floodplains and coastal deposits were built up on the margins of modern drainage systems and the coastline, respectively. 2.3.3 Depth to basement and major structural elements The OZ SEEBASE is a depth-to-basement structural model that provides a modelled depth to basement based on the shape and depth of major basement structures and the base of sedimentary sections. Depth to basement is modelled using a variety of data sources including magnetic and gravity data, digital elevation model (DEM) data and existing geological maps and publications (Geognostics, 2021). The SEEBASE map (Figure 2-8) shows the depth to basement and selected major structural elements within the Southern Gulf catchments using a colour gradient, with blue and green colours indicating deeper regions (to more than 10,000 mBSL) and orange and brown colours representing areas where the basement is present at shallower depths. Faults are represented by black dashed lines. Spatial changes in depth to basement highlight the major structural features beneath the Southern Gulf catchments and the map highlights that the Southern Gulf catchments are geologically and structurally very complex, as previously also described for some areas within the Southern Gulf catchments by Orr et al. (2020) and Buchanan et al. (2020). The depth to basement is shallow near the Gulf of Carpentaria and deepens significantly towards the south and west, where two major geological depocentres exist. The first depocentre is located between Doomadgee and Lawn Hill and hosts the stacked sequences of multiple sedimentary basins, including the Karumba Basin, Carpentaria Basin, South Nicholson Basin and Isa Superbasin (Figure 2-5). The largest basement depths (more than 10,000 m) in the Southern Gulf catchments are located between Lawn Hill and Camooweal. This depocentre hosts the thick sedimentary sequences of the Georgina and South Nicholson basins. Mid-sequence Cambrian volcanism and post-depositional neotectonic activity also affected the architecture of the basin. Figure 2-8 Major structural elements of the extended Southern Gulf catchments mMSL = metres relative to mean sea level. SEEBASE® image: Sourced from Geognostics (2021) 2.4 Hydrogeology 2.4.1 Aquifer types There are five main types of aquifers across the Southern Gulf catchments: (i) fractured and weathered rocks; (ii) fractured, fissured and karstic carbonate rocks of the Georgina Basin; (iii) extensive porous sedimentary sandstones of the GAB; (iv) porous sandstones of the Karumba Basin; and (v) surficial unconsolidated to consolidated alluvial sands and gravels (CSIRO, 2009a; 2009c; Smerdon et al., 2012), see Figure 2-9. These aquifer types occur in a variety of hydrogeological units (i.e. geological units hosting aquifers) across the catchments and host groundwater flow systems of varying scales. That is, some hydrogeological units host local-scale flow systems, such as the fractured and weathered rocks and alluvial sediments of the Lawn Hill Platform, South Nicholson Basin and Karumba Basin, respectively. Other hydrogeological units host intermediate to regional-scale flow systems, such as the fractured and karstic carbonate rocks of the Georgina Basin which hosts the Cambrian Limestone Aquifer (CLA), and the extensive porous sandstones of the Carpentaria Sub-basin of the GAB which hoists the Gilbert River Aquifer (GRA). Local-scale flow systems often have very short distances between recharge areas where water enters the groundwater system and discharge areas where water exits the system (i.e. a few hundred metres to a few kilometres). However, intermediate to regional-scale flow systems have much larger distances between recharge and discharge areas (i.e. several kilometres to tens or hundreds of kilometres). It is these larger groundwater flow systems that provide greater opportunities for groundwater development because they often: (i) store and transmit larger amounts of water, (ii) provide more opportunities for development away from existing users and groundwater-dependent ecosystems (GDEs), and (iii) have greater potential to coincide with larger areas of soils that may have potential for agricultural intensification. However, it must be noted that the local-scale flow systems of the fractured and weathered basement as well as alluvium and colluvium provide important groundwater supplies for stock and domestic use. Data and information for the different hydrogeological units and the groundwater flow systems they host vary significantly across the Southern Gulf catchments. For example, there are four ‘known’ hydrogeological units with intermediate to regional-scale groundwater systems that could potentially yield sufficient water for irrigated agriculture. These include the: (i) Gilbert River Formation in the Carpentaria Sub-basin of the GAB, (ii) Camooweal Dolostone, (iii) Wonarah Formation, and (iv) Thorntonia Limestone, all hosted within the Georgina Basin. There may also be other hydrogeological units that are currently poorly characterised but host local to intermediate- scale groundwater systems with potential to support small-scale localised irrigated agriculture opportunities (i.e. the Proterozoic basement rocks and the Bulimba Formation). However, data and information for these systems is very limited. Only part of the Gilbert River Formation, hosted in the GAB underlying approximately half of the north-east of the catchments, has been reasonably well characterised (Smerdon et al., 2012). Other hydrogeological units require more detailed information on their system conceptualisation, water quality, hydraulic properties and groundwater balance. Figure 2-9 Simplified regional aquifer-types of the Southern Gulf catchments Aquifer-types data source: Brodie et al. (2019) A map of different colors Description automatically generated 2.4.2 Hydrogeological units Great Artesian Basin –Carpentaria Basin The sequence of Jurassic to Cretaceous sediments of the Carpentaria Basin portion of the GAB were deposited on top of older Proterozoic geological basins, which give the general structure and shape of the GAB with the Jurassic to Cretaceous sediments becoming deeper and thicker to the north (Orr et al., 2020). The sequence of Jurassic to Cretaceous sediments in the Carpentaria Basin is comparable to that in the Eromanga Basin. The basal Jurassic–Cretaceous Eulo Queen Group and the overlying Gilbert River Formation sandstones are collectively equivalent to the Cadna-owie Formation – Hooray Sandstone Aquifer of the Eromanga Basin. The Wallumbilla Formation, Toolebuc Formation, Allaru Mudstone and Normanton Formation comprise the Rolling Downs Group in the Carpentaria Basin (Table 2-1). Table 2-1 Lithostratigraphy of the Carpentaria Basin (Great Artesian Basin) sediments. Adapted from Orr et al. (2020) and Smerdon et al. (2012) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au The Lower Cretaceous Gilbert River Formation is extensive onshore, while the Jurassic Eulo Queen Group extends northward offshore. Although the Eulo Queen Group has not been identified in the Southern Gulf Water Resource Assessment area, it could be present as thin isolated palaeovalley fill (Figure 2-10) (Buchanan et al., 2020; Orr et al., 2020). The Gilbert River Formation comprises sandstones with minor shale lenses and hosts the GRA which is considered the main GAB aquifer in the Southern Gulf catchments. The Gilbert River Formation commenced deposition in fluvial environments, followed by increasing marine influence. It is typically the deepest aquifer accessed for groundwater, with bores screened at a depth of 150 m to over 500 mBGS and thicknesses of up to 260 m (McConachie et al., 1997). Although bores with stratigraphic records are sparse, the onshore thickness of the Gilbert River Formation and its equivalents within the Southern Gulf catchments is likely to be thinner in most areas (Figure 2-5 and Figure 2-10) as suggested by refined three-dimensional geological models developed as part of Geoscience Australia’s project Assessing the Status of Groundwater in the GAB (Vizy and Rollet, 2023a; Wallace and Ransley, 2022). The hydraulic conductivity of the GRA is estimated to be 2 m/day (ranging from 0.1 to 10 m/day) (Klohn Crippen Berger, 2016). Groundwater within this aquifer is commonly under artesian pressure with flow rates initially recorded between 31 and 300 L/second, but these flows diminished substantially over time (Ingram, 1972). The formations of the Rolling Downs Group (Wallumbilla Formation, Toolebuc Formation, Allaru Mudstone and Normanton Formation) are generally considered aquitards or leaky aquitards, although the Normanton Formation is considered a partial aquifer (Buchanan et al., 2020). The Rolling Downs Group aquitard (comprised of the Wallumbilla Formation, Toolebuc Formation and Allaru Mudstone) is considered an effective aquitard in the Carpentaria Basin because of its lateral extent and continuity, considerable thickness (Figure 2-10) and the low permeability of the sediments. These were deposited in various marine settings with the Wallumbilla Formation being comprised of: (i) marine siltstone, claystone, glauconitic sandstone and silty limestone sequences (McConachie et al., 1997); (ii) the Toolebuc Formation of a thin unit of mostly limestone and carbonaceous shale (Cook et al., 2013); and (iii) the Allaru Mudstone of siltstone, claystone and minor fine-grained sandstone (McConachie et al., 1997). Intraformational polygonal faulting causes small vertical displacements in the order of tens of metres throughout the Rolling Downs Group aquitard (Buchanan et al., 2020). The Normanton Formation consists of sandstone and siltstone with minor glauconitic horizons (McConachie et al., 1997; Foley et al., 2022) of up to 300 m thickness with a partial aquifer characteristic where mean bore yields are 2 L/second (ranging from less than 1 to about 4 L/second) (Klohn Crippen Berger, 2016). The lithostratigraphy of the Carpentaria Basin sediments are given in Table 2-1. For further information on the Carpentaria Basin portion of the GAB see Buchanan et al. (2020); Orr et al. (2020); Smerdon et al. (2012); Vizy and Rollet (2023); Wallace and Ransley (2022) and Foley et al. (2022). Figure 2-10 Cross-section of the Carpentaria and Karumba basins in the Isa Geological and Bioregional Assessment (GBA) region (Buchanan et al., 2020) Gilbert River Formation –Carpentaria Basin The Gilbert River Formation occurs within the Jurassic to Cretaceous rocks and sediments in the north-east of the catchments and hosts the GRA (Figure 2-11). Bore depths in the Gilbert River Formation can be significant with respect to its potential for groundwater-based irrigation. Bore depths to intersect the extensive sandstone aquifer range between 125 and 750 mBGL (DOR, 2021a). Hydraulic conductivity values for the GRA in the northern Carpentaria and Laura basins range between 0.1 and 10 m/day, while transmissivities range between 4 and 570 m2/day (Horn et al., 1995; Klohn Crippen Berger, 2016). According to Ingram (1972), this aquifer often provides a well yield of 10 L/second from stock and domestic groundwater bores, which is consistent with findings across Cape York by Horn et al. (1995). A recent dataset compiled by the Australian Government Geological and Bioregional Assessment (GBA) project (n = 22), which partially overlies the GAB extent within the Southern Gulf catchments, indicated indicative yields of between 0.3 and 6.4 L/second (Buchanan et al., 2020). It is anticipated that with appropriately constructed large-diameter production bores, yields would likely exceed 10 L/second. The GRA is confined by the thick and laterally continuous aquitard of the Rolling Downs Group. In some places this results in artesian conditions, though not as pronounced as in other parts of Cape York (Horn et al., 1995; Taylor et al., 2018a). Groundwater salinity in the GRA can be fresh to brackish in places but still suitable for irrigation use (DOR, 2021a). Groundwater samples recently collected by Bardwell and Grey (2016) indicate a median salinity, represented by total dissolved solids (TDS) of 1324 mg/L, and its ionic composition is Na–HCO3–Cl and Na–Cl–HCO3, which are indicative of very long flow paths. An assessment of recharge processes and flow dynamics in the GAB by Raiber et al. (2022) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au identified the lack of environmental tracer data in the Carpentaria Basin as a key data gap within the GAB. Figure 2-11 Simplified regional hydrogeology of the Southern Gulf catchments Figure adapted from Figure SW-2 in CSIRO (2009a) A map of a country Description automatically generated Camooweal Dolostone – Georgina Basin The fractured and karstic carbonate rocks (Cambrian limestone, siltstone and dolostone) in the Georgina Basin host the Cambrian Limestone Aquifer (CLA) which also offers potential to support regional development based on existing hydrogeological data but remain poorly characterised. The Camooweal Dolostone (Cambrian dolostone), which occurs in the south-west of the catchments (Figure 2-11), is mostly comprised of dolostone and dolomitic limestone (Geoscience Australia and Australian Stratigraphy Commission, 2021; Matthews, 1992), and is the shallowest of the Georgina Basin’s more prospective hydrogeological units in the Southern Gulf catchments. At Carrara 1 (Figure 2-6), the thickness of the Camooweal Dolostone is more than 200 m and it has an approximate spatial extent for its outcropping/subcropping zone of about 18,277 km2 (CSIRO, 2009c; Taylor et al., 2021). The Camooweal Dolostone is one of two key hydrogeological units hosting the regional-scale CLA. Groundwater quality is mostly fresh to brackish with an electrical conductivity of 300 to 1500 μS/cm (Department of Environment, Parks and Water Security, 2021). The ionic composition of groundwater is predominantly Ca–Mg–HCO3 and therefore can cause scale build-up on groundwater bores (CSIRO, 2009c). Most bores intersecting the aquifer range in depth from 50 to 100 mBGL, though standing water levels (SWLs) can be deep (i.e. between 70 and 100 mBGL) ((Department of Environment, Parks and Water Security, 2021). Indicative bore yields derived from airlifting or submersible pumps at the time of drilling appear reasonable, ranging up to 14 L/second, though most bores yield 0.5 to 5 L/second (Matthews, 1992; Read, 2003). Data for the aquifers’ hydraulic properties is limited though is expected to be highly variable given the karstic nature of the aquifer. Investigations by Matthews (1992) and data collated by Read (2003) suggest transmissivities for the dolostone ranging between 290 and 1700 m2/day. Localised spring discharge and diffuse seepage from the Camooweal Dolostone supports dry-season baseflow to the Gregory River and its tributaries as well as Lawn Hill Creek in the Southern Gulf catchments (CSIRO, 2009c; Jolly and Tickell, 2011; Tickell, 2003). Both of these groundwater-dependent streams are listed as nationally significant wetlands (Environment Australia, 2001). Wonarah Formation – Georgina Basin The Wonarah Formation (Cambrian siltstone) of the Georgina Basin occurs in the south-west portion of the Southern Gulf catchments (Figure 2-11) and merges laterally into the Anthony Lagoon Formation (the Wiso Basin stratigraphic equivalent) (Tickell and Bruwer, 2017). The Wonarah Formation is predominantly composed of silty dolostone with interbedded (minor) dolomitic and calcareous mudstone and siliciclastic mudstone and has a maximum thickness of 244 m in holes drilled in its equivalent Anthony Lagoon Formation (Kruse and Dunster, 2013). Very little information exists for the Wonarah Formation in the Southern Gulf catchments. It therefore remains unclear if it exhibits the same characteristics as the Anthony Lagoon Formation in the Wiso Basin much further to the west in the NT. These formations are fractured, highly heterogeneous and partially karstic, with estimated transmissivities of between 13 and 8200 m2/day (Kruse and Dunster, 2013). Indicative hydrogeological data and anecdotal evidence for aquifers in the Wonarah Formation and equivalent comes from (Tickell, 2003). Tickell (2003) described the hydrogeological unit in the Barkly Tablelands as fractured and cavernous rocks assigned to both the Georgina and Wiso basins. Aquifers range in depth between 50 and 125 mBGL and standing water levels (SWLs) vary from 30 to 100 mBGL. Indicative bore-yield data based on airlifting at the time of drilling varies between 0.5 and 5 L/second. However, mud circulation losses during drilling were commonly reported, suggesting the presence of cavernous limestones, and potential for high-yielding aquifers. Groundwater quality was also described as fresh to brackish (i.e. TDS of 500 to 1500 mg/L) across most of the area investigated (Tickell, 2003). Thorntonia Limestone – Georgina Basin The Thorntonia Limestone (Cambrian limestone) occurs in the south-west of the catchments (Figure 2-11) and is comprised mostly of dolostone and dolomitic limestone (Geoscience Australia and Australian Stratigraphy Commission, 2021). It is overlain by the Camooweal Dolostone and is confined in parts by the Wonarah Formation. The limestone also hosts the CLA and occurs mostly in the NT but also outcrops/subcrops in far western Queensland (Figure 2-11). Its outcropping/subcropping zone has an approximate spatial extent of 6298 km2 and is up to approximately 100 m thick (CSIRO, 2009c; Taylor et al., 2021). Groundwater quality within the limestone aquifer is mostly fresh with an electrical conductivity of less than 500 μS/cm, making it suitable for irrigation use (Department of Environment, Parks and Water Security, 2021). The ionic composition of groundwater is predominantly Ca, Mg and HCO3 and therefore can result in scale build-up on groundwater bores (CSIRO, 2009c). Most bores intersecting the aquifer range in depth from 50 to 100 mBGL, and SWLs can be relatively deep (i.e. between 50 and 100 mBGL) ((Department of Environment, Parks and Water Security, 2021). Indicative bore yields derived from airlifting or submersible pumps at the time of drilling appear reasonable, with most bores yielding 0.5 to 5 L/second. There are currently no known digitised and interpreted pump test data for the aquifer and therefore no data on aquifer properties. However, aquifer properties are expected to be highly variable but also exhibit high permeability in places due to the karstic nature of the aquifer and extensive network of interconnected solution cavities (CSIRO, 2009c). Localised spring discharge and diffuse seepage from the Thorntonia Limestone also supports dry-season baseflow to the nationally important Gregory River and its tributaries as well as Lawn Hill Creek in places across the Southern Gulf catchments (CSIRO, 2009c; Jolly and Tickell, 2011; Tickell, 2003). Other hydrogeological units There is some useful information available for two other hydrogeological units hosted in the Proterozoic fractured and weathered rocks comprising the regional basement and the Cenozoic shallow sediments of the Karumba Basin, mostly associated with alluvial and colluvial deposits. Proterozoic rocks CSIRO (2009c) described the Proterozoic rocks trending north-west to south-east across the central part of the Southern Gulf catchments (Figure 2-11) as not being a feasible groundwater resource. However, as pointed out by (McEniery, 1980), until the construction of Lake Moondarra Dam, Mount Isa town water supply was sourced from fractured shales. These shales occur at depths ranging from 60 to 80 mBGL and yield between 5 and 10 L/second. In addition, about 50% of the bores with stratigraphic information in the Southern Gulf catchments correspond to Proterozoic units composed of unassigned granites or attributed to one of the following geological units: Lawn Hill, Corella, Paradise Creek, Surprise Creek, Toole Creek Volcanics, Lady Loretta and Gunpowder Creek. According to the data compiled by (Bardwell and Grey, 2016), a significant spatial variability in groundwater chemistry is found among the Proterozoic aquifers, with dominant ions varying from Mg–Ca–HCO3 to Na–SO4. Such characteristics may be an indication of localised recharge and interactions of groundwater with variable minerally rich rocks. A cluster of 22 perennial springs assigned to the Lawn Hill Formation is present in the Southern Gulf catchments (Figure 2-11) according to the Queensland Springs Database (Queensland Government, 2021). This group of springs is located at the centre of the catchment of the Nicholson River, in relatively near proximity to 14 groundwater bores also assigned to source water from the Lawn Hill Formation. Such evidence suggests that the Proterozoic fracture rocks host a dynamic and complex hydrogeological system that supports dry-season baseflow to Lawn Hill Creek. Bulimba Formation Consisting of lacustrine, fluvial and, to a lesser extent, shallow shelf marine sediments, the widespread Cenozoic sediments of the Karumba Basin unconformably overlie the Carpentaria Sub- basin of the GAB (Bradshaw et al., 2009) in the north-east of the Southern Gulf catchments (Figure 2-11). Within the Southern Gulf catchments, Cenozoic sediments are expected to reach a maximum thickness of 40 m (north of Doomadgee – see Figure 2-12), with the most productive aquifer corresponding to the basal section of the Bulimba Formation, composed mostly of fine- grained quartzose sediments (Herbert, 2000; Radke et al., 2012). Lithological properties of the Bulimba Formation are highly variable, from shale to sandy ferricrete, resulting in variable hydraulic properties and groundwater yields (i.e. 0.25 to 4.5 L/second) (DOR, 2021a). Nevertheless, at the central area of the catchment of the Nicholson River, where the alluvial deposits are wider (downstream of Doomadgee) (Figure 2-12), the Bulimba Formation corresponds to the most accessed aquifer from the Karumba Basin (Buchanan et al., 2020). Its hydraulic conductivity ranges from 150 to 300 m/day with specific yield of 0.1 (Smerdon et al., 2012). Figure 2-12 Thickness of Cenozoic sediments in the Southern Gulf catchments Data source: Vizy and Rollet (2023) 2.5 Surface water hydrology The Southern Gulf catchments consist of the contributing area of various rivers and streams that discharge into the southern Gulf of Carpentaria. The most substantial of these are the Leichhardt, Gregory and Nicholson rivers. The catchments of these rivers, plus those on the Wellesley Islands in the Gulf of Carpentaria, Settlement Creek, Morning Inlet, and numerous small coastal creeks, have a total area of 108,200 km2. The Leichhardt catchment has an area of 33,400 km2, and the river itself extends approximately 550 km from the river mouth to Mount Isa in the south of the catchments. On the Leichhardt River, brackish salinity is likely to extend as far up as Leichhardt Falls, 113 km from the mouth. The Leichhardt River is ephemeral and flows less than half the time. Lake Julius and Lake Moondarra are two large storages (approximately 107 GL each) in the upper Leichhardt River, used for town and industrial water supply near Mount Isa. Gauges (Figure 2-13) on the Gregory and O’Shannassy rivers indicate perennial flow (and almost perennial flow in Lawn Hill Creek), with discharge from the Thorntonia Limestone hydrogeological unit maintaining dry-season baseflow. Elsewhere in the Nicholson catchment the streams are ephemeral. There is a distinct rainfall gradient across the Southern Gulf catchments, with the highest rainfall along the coast in the north at over 900 mm/year and the lowest in the south furthest inland at below 500 mm/year. The majority of annual rainfall is in the three months from January to March. The runoff and streamflow follow the same pattern as the rainfall with higher runoff generated towards the coast during the wet season. The median and mean annual discharges from all Southern Gulf catchments are 4961 and 6759 GL/year, respectively, with the majority of discharge being from the Leichhardt and Nicholson rivers (Figure 2-13). The pronounced difference between the mean and median is due to the mean being ‘biased’ by a number of very high flow years. These high flow years result in extensive flooding, mostly in the downstream reaches and especially in between the Nicholson, Gregory and Leichhardt rivers. Figure 2-13 Streamflow observation data availability and median annual streamflow (50% exceedance) under Scenario A in the Southern Gulf catchments For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 2.6 Water-dependent ecosystems The freshwater sections of the Southern Gulf catchments include diverse habitats such as persistent and ephemeral rivers, wetlands, floodplains and GDEs. The diversity and complexity of habitats, and the connections between habitats within a catchment, are vital for providing the range of habitats needed to support both aquatic and terrestrial biota (Schofield et al., 2018). In the wet season, flooding connects rivers to floodplains. The exchange of water from the river across the floodplain supports higher levels of primary and secondary productivity than compared to surrounding areas with less frequent inundation (Pettit et al., 2011). Infiltration of water into the soil during the wet season and along persistent streams often enables riparian habitats to form an important interface between the aquatic and terrestrial environment. While riparian habitats often occupy a relatively small proportion of the catchments, they frequently have a higher species richness and abundance of individuals than surrounding habitats (Pettit et al., 2011; Xiang et al., 2016). In the dry season, biodiversity is supported by the perennial rivers and creeks (Gregory River, O’Shannassy River and Lawn Hill Creek; Figure 2-14), permanent lakes, and in-channel waterholes. These water sources that persist through the dry season become increasingly important as the season progresses, as they provide important refuge habitat for species and enable recolonisation into surrounding habitats upon the return of larger flows (Hermoso et al., 2013). These water sources provide habitat for water-dependent species including fish, sawfish and turtles, as well as providing a source of water for other species more broadly within the landscape (McJannet et al., 2014; Waltham et al., 2013). Figure 2-14 presents the current state of knowledge about the type and distribution of GDEs and surface water dependent ecosystems in Southern Gulf catchments. Terrestrial GDEs were mapped based on occurrences of vegetation species that are known only to grow where they have access to groundwater (known GDE: Eucalyptus camaldulensis, Melaleuca argentea, Barringtonia acutangula) and mapped persistence of vegetation throughout seasons (Castellazzi et al., 2024) based on remote sensing analysis and ground-truthing. Many other vegetation species present in Southern Gulf catchments may also be dependent on access to groundwater, however they are unconfirmed and therefore not included in Figure 2-14. Aquatic GDEs were mapped using existing datasets: groundwater-fed springs, persistent lakes, rivers, wetlands and waterholes. Known subterranean GDEs are hosted in karstic aquifer and cave systems. Mapped ecosystems that require surface water inundation include rivers, reservoirs, Directory of Important Wetlands in Australia wetlands, swamps, land subject to inundation and vegetation species that are known to require surface water inundation (Eucalyptus camaldulensis, paperbark swamp species, monsoon vine forest species that are diagnostic of flood inundation; Merrin et al., 2023). Figure 2-14 Water-dependent ecosystems in Southern Gulf catchments including terrestrial, aquatic and subsurface environments that require surface water inundation and/or access to groundwater GDE = groundwater-dependent ecosystem. Data sources: Nationally important wetlands – Environment Australia (2001); known and high confidence GDEs – Doody et al. (2017); high probability groundwater-dependent vegetation – Castellanzzi et al. (2024); Atlas of Living Australia (2003), Department of Environment Parks and Water Security (2000a); NVIS Technical Working Group (2017) A map of the north america Description automatically generated Part IIMethods Chapter2 Study area|39Leichhardt FallsPhoto:CSIRO 3 Regional desktop and modelling assessment of the Southern Gulf catchments Section 3 of the report describes the methods used to conduct a regional desktop assessment of available hydrogeological data across the Southern Gulf catchments to: (i) provide an overview of available data across the catchments, (ii) collate data for input into regional-scale recharge modelling and mapping of potential groundwater discharge areas, and (iii) to generate baseline datasets for use in more detailed field, desktop and modelling investigations of the Cambrian Limestone Aquifer (CLA) and Gilbert River Aquifer (GRA). 3.1 Regional geological and hydrogeological desktop assessment A regional hydrogeological desktop study was undertaken across the catchments of the Southern Gulf rivers, that is Settlement Creek, Gregory–Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups, to evaluate the data availability for all hydrogeological units. This study provided an aquifer attribution to publicly available groundwater data available through: (i) the NT Government Department of Environment, Parks and Water Security web mapping tool (Department of Environment, Parks and Water Security, 2021) for catchments in the NT, and (ii) the Queensland Groundwater Database (DOR, 2021a), as well as bore reports available through Queensland Globe (DOR, 2021b). Aquifer-specific hydrogeological data included water levels and quality, water chemistry, bore yield and aquifer properties. These different data sources were collated and summarised along with data published in existing literature to evaluate the amount of available information for each of the different hydrogeological units in the Southern Gulf catchments. The desktop geological and hydrogeological assessment was designed to underpin targeted field investigations to address key knowledge gaps associated with system conceptualisation, water quality, aquifer properties and groundwater balances for the most prospective hydrogeological units. However, due to access issues, no geological and hydrogeological field campaigns could be conducted. Field activities that were planned during the earlier stages of the project included: (i) conducting drilling or ground-based geophysical surveys to better characterise aquifer extents and geometries, saturated thickness and aquifer properties; (ii) deploying monitoring equipment to candidate groundwater bores to implement hydrological monitoring; (iii) conducting a groundwater hydrochemistry and environmental tracer sampling program; and (iv) characterising areas with culturally and ecologically significant groundwater-dependent ecosystems (GDEs) where groundwater–surface water interaction investigations may be required. The aim of the planned monitoring and sampling programs was to assist in further characterising, conceptualising and quantifying groundwater flow processes (i.e. recharge, throughflow and discharge) and where possible, seek to evaluate the potential for future development. Nevertheless, the desktop geological and hydrogeological assessment was used to provide baseline datasets to conduct detailed desktop and modelling investigations outlined in Section 4. 3.1.1 Geological framework A 1:1,000,000 surface geological map (Raymond et al., 2012) was used as the basis for the geological assessment of the Southern Gulf catchments. This map was further simplified to achieve a smaller number of geological classes for representations on maps, cross-sections and for aquifer attributions. 3.1.2 Pre-existing three-dimensional geological models Pre-existing three-dimensional geological models and other geoscientific datasets were compiled, including: •Geoscience Australia Eastern Australia three-dimensional geological and hydrogeologicalsurfaces update, Version 2.0 (Vizy and Rollet, 2023a) •Great Artesian Basin Vshale modelling, Version 1.0 ‒ Data package of Vshale modelling based onthe Eastern Australia three-dimensional geological and hydrogeological surfaces update, Version2.0 (Vizy and Rollet, 2023b) •Preliminary three-dimensional geological model of the Georgina‒southern McArthur‒SouthNicholson basins ‒ Data package of three-dimensional modelling geological surfaces andisochores, Version 1.0 (Vizy and Rollet, 2024) •Regional stratigraphic correlation transects across the Great Artesian, Lake Eyre basins andUpper Darling Floodplain region (part 2): Infilling data and knowledge gaps (Norton and Rollet, 2023) •Digital elevation model (DEM) data (Shuttle Radar Topography Mission (SRTM)-derived 1 SecondDigital Elevation Models Version 1.0. Record 1 (Gallant et al., 2011) •Georgina Basin Geoscience Data Package (Carr et al., 2016). 3.2 Mineralogical analysis of core samples Twenty-two core samples were collected for mineralogical analysis from the Dobbyn-1 exploration well from the Queensland core library (Exploration Data Centre, Zillmere, Qld; Figure 3-1). Multiple samples were collected for each intersected stratigraphic at this well to investigate the mineralogical variability within each formation. Understanding the mineralogical variability is important as the mineralogical composition of a sedimentary rocks has a significant influence on groundwater flow within the rock. For example, sedimentary rocks with a high abundance of quartz and a low abundance of clay minerals such as kaolinite and smectite typically form productive aquifers. In contrast, sedimentary rocks with low quartz content and a dominance of clay minerals typically act as partial aquifers or aquitards. Core samples were analysed for bulk mineralogy by X-ray powder diffraction (XRD) and elemental composition analysis by X-ray fluorescence spectroscopy (XRF) analysis at the CSIRO Mineral Resources XRD and XRF laboratory in Adelaide. Figure 3-1 Queensland core library at the Exploration Data Centre, Zillmere, Qld Data source: CSIRO For bulk quantitative XRD phase analysis, the sedimentary rocks were crushed in a ring mill and approximately 1.5 g of each pulverised powder sample was ground in ethanol in a McCrone micronizing mill. The resulting slurry was oven dried at 60°C then thoroughly mixed in an agate mortar and pestle before being lightly back pressed into a stainless-steel sample holder for presentation to the X-ray beam. The sample was then Ca exchanged in CaCl2 solution to ensure that the phases that have exchangeable cations (e.g., smectite) are in a divalent monoionic form. The Ca-exchanged sample was lightly back pressed into a stainless-steel sample holder for presentation to the X-ray beam. Approximately 0.8 g of each oven dried (105°C) bulk sample was accurately weighed with 8 g of 12:22 lithium borate flux. The mixtures were heated to 1050°C in a Pt/Au crucible for 20 minutes to completely dissolve the sample then poured into a 40 mm Pt/Au mould heated to a similar temperature. The melt was cooled rapidly over a compressed air stream and the resulting glass disks were analysed under vacuum using a Bruker S8 Tiger wavelength dispersive X-ray fluorescence spectrometer. Phase identification was performed using PANalytical Highscore Plus© software (V4.8) which interfaces with the International Centre for Diffraction Data (ICDD) PDF 4+ 2023 database. Quantitative phase analysis (QPA) was carried out via the Rietveld method using the TOPAS V7 software. The QPA results obtained are the relative weight percentages (wt%) of crystalline phases. 3.3 Groundwater bore data 3.3.1 Lithological and stratigraphic logs Bore logs (also commonly known as driller’s logs) or well logs and stratigraphic logs are the main sources of data used to reliably assign screened intervals (corresponding to the interval where the casing is slotted to allow intake from the aquifer) to discrete stratigraphic units. Lithological logs 42 | Characterising groundwater resources in the Southern Gulf catchments describe the composition of the strata, whereas stratigraphic logs assign depth intervals to discrete stratigraphic units. Groundwater bore logs and associated groundwater-level measurements are also essential for constructing potentiometric surface maps that show groundwater flow direction and for characterising spatial and/or temporal water chemistry patterns within aquifers. A typical bore/well log includes the following information: • a registration number that uniquely identifies the bore • a location (easting and northing, or longitude and latitude) • the total drilled depth • bore/well type (e.g. groundwater, exploration, appraisal or production) • an elevation of the natural ground surface or the top of the casing • the depth interval where the groundwater bore is open/perforated to the aquifer (screened interval) • lithological and stratigraphic descriptions with their associated depth intervals. The major bore/well log data sources used in this assessment are: • groundwater bores, sourced from the Queensland Department of Regional Development, Manufacturing and Water groundwater database, and from the NT Land Information System (NTLIS) • selected deep stratigraphic wells, sourced from publications and data access portals (e.g. Carrara 1) (Geoscience Australia Portal, 2023). 3.3.2 Queensland Groundwater bore data for Queensland were downloaded from the Queensland Groundwater Database (Department of Regional Development, Manufacturing and Water, 2023). For Queensland, bore location, bore construction data and bore stratigraphy were obtained directly from the Queensland groundwater database. Multiple tables were linked and queried within an Access database. 3.3.3 Northern Territory Groundwater bore data were downloaded as portable document formats (PDFs) for each bore from the NTLIS. Lithological logs and associated bore information, which generally only existed as hard copies, were digitised and extensive data quality checks were performed. The lithological logs were then interrogated in three-dimensional space using GoCAD-SKUA three-dimensional geological modelling software, subjected to further data quality checks, and simplified to reduce the number of lithological classes for representation in the hydrogeological and hydrochemical assessment. For some of the groundwater bores in the NT, the bore records did not contain any lithological information and the majority of bore records here did not include stratigraphic descriptions. Where available, other groundwater bore information (e.g. yields and static water level) were also manually digitised for each bore. Some common problems with lithological and stratigraphic descriptions in databases are: •no lithological and/or stratigraphic information available for many bores or wells •omission of important information such as the colour of sandstones •use of incorrect geological terms •no depth information •misspellings. Consequently, all well log data were subjected to extensive data quality checks prior to use for hydrogeological or hydrochemical applications. All information was standardised (e.g. depth intervals were converted from feet to metres and yields were converted from gallons to litres per time unit). 3.4 Aquifer data 3.4.1 Simplification of stratigraphic classes The 1:1 million surface geology map (Raymond et al., 2012) of the extended Southern Gulf catchments area included more than 220 stratigraphic classes. As highlighted in Section 3.3.1, many groundwater bores in Queensland have stratigraphic data or aquifer descriptions, whereas most bores in the NT only have lithological logs available (with <10% having stratigraphic data within the extended Southern Gulf catchments). Many of the stratigraphic units in the Southern Gulf catchments are composed of variable sequences of limestone, dolostone, sandstone, mudstone, siltstone, igneous rocks or other rock types and there are multiple formations with similar characteristics. For example, within the Georgina Basin, there are multiple stratigraphic formations composed of limestone, dolostones or other carbonate rocks that are difficult to differentiate based on lithological logs. As a result, rock types are not necessarily characteristic markers of different formations, and the assignment of depth horizons and screened intervals to unique stratigraphic classes based on lithological logs can be challenging and is subjected to high uncertainties. To overcome this limitation, an attempt to assign depth horizons and screened intervals to broad stratigraphic classes based on lithology was done based on geological knowledge and the expert scientific judgment of the Assessment team. Due to the complexity of the data and the geological setting and to achieve a manageable number of stratigraphic classes that allows representation of groundwater data on hydrogeological maps, the surface geology dataset was simplified from more than 220 stratigraphic classes into a smaller subset of lithological descriptions (Figure 3-2). Figure 3-2 Simplified regional geology of the Southern Gulf catchments This map does not represent outcropping areas of all hydrogeological units: the blanket of surficial Cretaceous to Quaternary regolith sediments has been removed to highlight the spatial extent of various regional hydrogeological units in the subsurface. Geology data sources adapted from: Department of Industry, Tourism and Trade (2014) and Department of Environment, Parks and Water Security (2008) and Raymond et al. (2012) Spring data sources: Department of Environment, Parks and Water Security (2013); the Groundwater Dependent Ecosystems Atlas (Bureau of Meteorology, 2017) and Department of Environment, Science and Innovation (2021) After simplification, a smaller number of main classes remained and were taken forward for the hydrogeological and hydrogeochemical assessment (Figure 3-2). These main classes include: •Quaternary alluvium •Cenozoic fluvial •Cretaceous sandstone, mudstone and conglomerate •Cambrian dolostone and limestone •Cambrian sandstone and siltstone •Igneous rocks •Proterozoic dolostone, sandstone, mudstone, quartzite and conglomerates 3.4.2 Aquifer attribution A key challenge in hydrogeological investigations is the availability of construction and aquifer information for groundwater bores. Bore construction and reporting standards have evolved over time, and therefore the quality and completeness of bore records held in state groundwater databases (Department of Regional Development, Manufacturing and Water, 2023) is variable. As described in Section 3.3.3, groundwater bore data were downloaded as PDFs for each bore from the NTLIS and digitised. In some cases, there is incomplete or limited construction information available and inconsistent information about source aquifers. Furthermore, where stratigraphic data do exist, they might either be incorrect and/or only cover part of the bore depth profile; the depth of bores is often unknown, as are the top and/or bottom of the screened interval. In this study, bore casing construction information (e.g. depth of screened interval) was used together with other geological and hydrogeological data (e.g. geological maps or stratigraphic logs from nearby wells) to assess aquifer membership. Aquifers at the screened interval of bores were classified according to the simplified aquifer classes presented in Section 3.4.1; their outcropping areas are shown on Figure 3-2. 3.5 Groundwater levels Groundwater levels in the Queensland part of the Southern Gulf catchments were obtained from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023). Depth to groundwater is an important attribute for assessing local and regional trends in water levels, identifying recharge and discharge areas and understanding the spatial changes in depths from which water would have to be pumped to the surface for use. The primary source of static groundwater-level data for groundwater bores across the Southern Gulf catchments was the NT Government (NTG) Open Data Portal, which provides a direct link to the digital spatial groundwater database for the entire NT (Department of Environment Parks and Water Security, 2014). In addition to digital data from the groundwater database, drilling records (hand-written, typed or digital) were also accessed via Natural Resources (NR) Maps (Department of Environment Parks and Water Security, 2021) to undertake the aquifer attributions described in Section 3.1. These drilling records also include observations of groundwater level, groundwater salinity and bore yield from testing undertaken when the bores were first constructed and installed. Depending on how old the drilling records are, some data have not yet been digitised and added to the territory-wide groundwater database. Therefore, where groundwater-level observations were absent in the territory-wide database, observations from the drilling records were obtained. Static groundwater-level observations were then linked to the aquifer attribution dataset in Section 3.1 and then symbolised and mapped in classes in ArcGIS by aquifer where aquifer information was available. 3.6 Groundwater chemistry and salinity Groundwater salinity and groundwater chemistry are important attributes for assessing the suitability of groundwater for different uses (i.e. domestic use, stock water, irrigation water or industrial applications) as well as identifying local and regional trends in spatial changes in salinity. Salinity data in the form of total dissolved solids (TDS) were collated from the territory-wide groundwater database accessed via the NTG Data Portal (Department of Environment, Parks and Water Security, 2014) and from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023). Due to the limited spatial coverage of available data within the Southern Gulf catchments, groundwater monitoring bores were selected both within and outside the catchments to provide a better representation of regional groundwater chemistry. Prior to mapping and evaluating the salinity data a review of the validity and quality of the analytical data was undertaken by calculating the charge balance error (CBE) using Geochemists Workbench software. Ideally, in hydrochemical studies, only samples with a CBE of ±5% should be used to interpret hydrochemical patterns and infer hydrogeological processes. However, in data- sparse regions the CBE of samples considered in an assessment can be widened to include a larger number of hydrochemical records. In this study only water samples with a CBE ±15% were used. A relatively high CBE was chosen due to the sparsity of data in some parts of the Southern Gulf catchments. Furthermore, an initial screening showed that some groundwaters within the Southern Gulf catchments are extremely fresh (TDS <50 mg/L), note that low-salinity groundwaters often have a significant CBE. The salinity data were then linked to the aquifer attribution dataset discussed in Section 3.4.2 and then symbolised and mapped in ArcGIS by aquifer where aquifer information was available and mapped in classes suitable for different water uses including stock water and irrigation. Hierarchical cluster analysis (HCA) was conducted using Statgraphics (version 19) Centurion software (Manugistics Inc., United States) using the methodology described in Martinez et al. (2015) and Raiber et al. (2019). HCA is a method that allows the incorporation of any user-defined combination of chemical and physical constituents, including non-numerical parameters, and allows the incorporation of any number of variables (Güler et al., 2002). In this study, ten variables were selected (Ca, Mg, Na, K, HCO3, Cl, F, SO4, pH and TDS concentrations) as they were measured across most sites. Temperature was not included as no field temperature measurements were available for many bores. HCA only considers complete cases where all selected input parameters have a measured value. In total, 1280 hydrochemical records from 742 groundwater bores (including multiple hydrochemical records for some bores) were included in the study. Prior to the multivariate statistical analysis, all variables except for pH were log transformed to ensure that each variable more closely follows a normal distribution. Two different linkage rules were used in HCA in this study: (i) the nearest neighbour linkage rule was used to identify monitoring sites with dissimilar hydrochemical signature to other sites (i.e. outliers), which could then be placed as residuals in a separate group; and (ii) Ward’s linkage rule, which generates distinct clusters based on an analysis of variance, was then used to group all non-residuals into separate clusters, where each site in a cluster is more similar to other sites in the same cluster than to any site from a different cluster. The square of the Euclidean distance (E) was used in HCA as the measure of similarity performed over all n variables included in HCA. Residuals were subsequently separated and HCA using Ward’s algorithm was conducted on the remaining complete samples where all input variables were measured. During the clustering procedure, HCA first considers each observation separately, and then combines the two observations that are closest together to form a new group. After re-computing the distance between the groups, the two groups then closest together are combined, and this process is repeated until only one group remains. Resulting from this procedure is a dendrogram (Figure 3-3). Piper and Schoeller plots for the median values of each cluster were constructed using AcQA software (Rockware). Figure 3-3 Dendrogram of groundwater chemistry data showing separation threshold for five clusters. Each vertical blue line corresponds to one groundwater chemistry record 3.7 Bore yield Bore-yield data is an attribute that provides an indication of the influence of the physical properties of the rocks and sediments of different aquifers on controlling the storage and flow of groundwater at different locations. It should be noted that long-term (12 to 48 hour) pump tests on production bores provide the best indication of bore yield at a certain location in an aquifer. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au However, the majority of bore-yield data in most locations within the extended Southern Gulf catchments is obtained by short-term (i.e. a few hours) discharge testing by air lifting or a submersible pump on small-diameter (i.e. <150 mm) stock and investigation bores. In most cases, the bore construction and size of the air compressor or submersible pump limit the yield from these bores. In addition, for artesian bores, the free-flowing yields are often recorded. Nevertheless, despite these limitations, the available data can provide a useful indication of aquifer potential. Bore-yield data were collated from the territory-wide groundwater database accessed via the NTG Data Portal (Department of Environment, Parks and Water Security, 2014a) and the Queensland Groundwater Database (Department of Regional Development, Manufacturing and Water, 2023). Within the NT, in addition to digital data from the NT groundwater database, drilling records (hand-written, typed or digital) were also accessed via NR Maps (Department of Environment Parks and Water Security, 2021) to undertake the aquifer attributions described in Section 3.4.2. Drilling records were reviewed for bore-yield data and where the territory-wide database did not have a bore-yield observation for a specific bore site, data from the drilling records were used where available. Bore-yield data were then linked to the aquifer attribution dataset in Section 3.4.2 and then symbolised and mapped in different classes in ArcGIS by aquifer, where aquifer information was available. 3.8 Recharge estimation In this study, recharge was estimated using the chloride mass balance (CMB). This approach is an evolution of the method used in previous Northern Australia Water Resource Assessment project (Taylor et al., 2018b; Taylor et al., 2018c; Turnadge et al., 2018), the Bioregional Assessments (Crosbie et al., 2018) and Geological and Bioregional Assessment (GBA) projects (Crosbie and Rachakonda, 2021) and most recently in the Sydney region (Wilkins et al., 2022b) and Great Artesian Basin (GAB) (Crosbie et al., 2022). It includes four steps (Figure 3-4): • Estimating recharge using the CMB method at a point scale. This includes estimating the chloride in rainfall, runoff and groundwater. • Regression kriging of the point estimates of recharge. This involves upscaling recharge using regression equations, kriging the residuals of the point-scale recharge and aggregating the regression rasters and residual rasters to form the regression-kriging upscaled recharge rasters. • Evaluating the uncertainty in the spatial recharge estimates using 1000 replicates of the two processes above. • Constraining the uncertainty using observations of baseflow at the low end and excess water at the high end. These four steps are described in detail in the following sub-sections. Results are found in Section 5.5. Figure 3-4 Schematic diagram of the methods used showing the calculation of the point recharge, the upscaling using regression kriging and the reporting of the results at the scale of the region and surface geology group The area under investigation for recharge using the CMB method is substantially larger than the Southern Gulf catchments. It includes both the catchments of the Roper and Southern Gulf rivers as they are linked by the Georgina Basin, the gulf catchments in between and the northern-flowing catchments that receive groundwater flow from the Dook Creek Formation (Figure 3-5). For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Stochastically generate chloride deposition surface (single realisation for all points) Stochastically generate chloride in runoff(each point individually generated) Stochastically generate chloride in groundwater(each point individually generated) Calculate point rechargeDevelop regression equation using covariatesCalculate recharge residualsKrige residuals to regular gridUpscale recharge to regular grid using covariatesAdd recharge and residuals gridRaster Stack of 1000 replicates of upscaled recharge estimatesConstrain uncertainty using baseflow and excess waterSummary of replicates as 5th, 50thand 95thpercentilesRepeat 1000 timesResults aggregated by aquifer and surface geology Figure 3-5 Region used for estimating recharge using the chloride mass balance method A map of the united states Description automatically generated 3.8.1 Point-scale chloride mass balance The CMB method (Anderson, 1945) is the most widely used approach for estimating recharge, both globally (Scanlon et al., 2006) and in Australia (Crosbie et al., 2010). It is popular because it is robust over many climate zones and is cost effective, requiring only analyses of chloride in groundwater and rainfall. It is relatively insensitive to the mechanism of recharge, whether that be diffuse recharge through the soil matrix, bypass flow through macropores or even captured surface water in sinkholes (Alcalá et al., 2011; Bazuhair and Wood, 1996). At its simplest, the CMB method can be used to estimate recharge (R, measured in millimetres per year) knowing only the chloride deposition from rainfall (D, measured in kilograms per hectare per year) and the chloride concentration of the groundwater (Clgw, measured in milligrams per litre): 𝑅𝑅=100 𝐷𝐷 𝐶𝐶𝑙𝑙gw (1) This works because the chloride deposited on the land surface from the atmosphere is excluded from evaporation and transpiration and therefore the chloride becomes concentrated as it travels through the soil profile to the groundwater. This evapoconcentration allows estimation of the recharge rate. Wood (1999) listed the important assumptions behind estimating recharge using the CMB method as: 1.Chloride in groundwater originates from rainfall on the aquifer and not from flow fromunderlying or overlying aquifers. 2.Chloride is conservative in the system. 3.Steady-state conditions are assumed in that the fluxes of chloride and water have not changedover time. 4.There is no recycling of chloride within the aquifer. These assumptions are discussed in the following paragraphs. For most of the study area, the chloride in groundwater is due to rainfall. In some alluvial areas there is additional chloride added to the groundwater system through losing streams and recharge due to overbank flooding. Chloride in groundwater observations from these areas needs to be identified and removed from the analysis. In confined aquifers there can be some mixing of groundwater from layers above and below, so to avoid these complications only groundwater chloride observations from outcropping areas are in the analysis. Chloride is conservative in the system when there are no sources or sinks in the system (i.e. all chloride is derived from rainfall and it does not react with minerals). There is chloride export from the system through runoff, and this is accounted for by incorporating runoff into Equation (1). The steady-state assumption is difficult to meet in areas that have undergone recent land use change (Cartwright et al., 2007) such as the clearing of native vegetation for agriculture. There has been limited land clearing in the areas under investigation and so the assumption is made that groundwater observations relate to the current land use. To avoid ‘old’ water and only sample ‘young’ water, chloride in groundwater observations from outcrop areas and in bores that are shallower than 100 m are used. Recycling of chloride occurs when groundwater is evaporated or transpired and then returned to the groundwater system. Recycling can occur in groundwater discharge areas (Bazuhair and Wood, 1996), flow through lakes (Howcroft et al., 2017) and through irrigation with groundwater (Wood and Sanford, 1995). These areas can be easily identified as having actual evapotranspiration greater than rainfall and so any chloride in groundwater observations from these areas can be excluded. Chloride in rainfall The chloride deposition rate (kilograms per hectare per year) has been obtained from a national gridded product at a 5 km resolution (Wilkins et al., 2022a). Because of the sparsity of observational data concerning chloride in rainfall, the uncertainty in the result is high in certain regions. Figure 3-6 shows the chloride deposition rate in the study area. The mean (μ), standard deviation (σ) and skewness (g) from 1000 replicate models are used to provide a Pearson Type III distribution of chloride deposition at each point location in the study area (Figure 3-6). The deposition according to the Pearson Type III distribution is given as (Pilgrim, 1987): 𝐷𝐷=𝜇𝜇+𝐾𝐾𝑌𝑌.𝜎𝜎 (2) where KY is a frequency factor calculated from g and a standard normal deviate (z): 𝐾𝐾𝑌𝑌= 2 𝑔𝑔 􀵤􁉄􁉀𝑧𝑧− 𝑔𝑔 6 􁉁 𝑔𝑔 6+1􁉅 3−1􀵨 (3) The standard normal deviate is generated stochastically for each replicate and used to generate the chloride deposition rate for input into the recharge estimation at each point location. The same standard normal deviate is used for each bore location in a replicate to maintain spatial consistency in the recharge estimates. The same process is repeated 1000 times to generate the inputs for the 1000 replicates. Figure 3-6 The (a) mean, (b) standard deviation and (c) skewness of chloride deposition across the study area (Wilkins et al., 2022a). The blue squares indicate points where chloride deposition has been measured For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Chloride in runoff The chloride exported in runoff can be accounted for by modifying Equation (1) by assuming that the chloride exported in surface runoff is proportional to the runoff coefficient. The runoff coefficient (RC), which is the proportion of rainfall that becomes streamflow, can be obtained from the output of the Australian Water Resource Assessment (AWRA) model (Vaze et al., 2013) on a gridded basis and therefore can then be used to estimate RC at each point within the study area (Figure 3-7). Figure 3-7 Runoff coefficient across the study area For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au To account for runoff, Equation (1) can be modified to: 𝑅𝑅=100𝐷𝐷(1−𝛼𝛼.𝑅𝑅𝑅𝑅) 𝐶𝐶𝐶𝐶gw (4) Here α is a scale factor, which is included because large rainfall events that result in large runoff volumes are often below average in chloride concentration, and therefore do not reduce chloride deposition proportionally. For this study, α is sampled stochastically from a uniform distribution with a value between 0.33 and 0.66 (Crosbie et al., 2018) during the uncertainty estimation (the second step of Figure 3-4). The value of α is equal for each bore in a replicate to maintain spatial consistency but is sampled individually for each replicate. Chloride in groundwater Measurements of chloride in groundwater were sourced from two datasets: 1.the NT gov website (Department of Environment, Parks and Water Security (2014) 2.the Bureau of Meteorology’s (BoM) National Groundwater Information System (NGIS). There were too many individual observations of chloride in groundwater to enable quality assurance and quality control to be undertaken on each observation, so an automated procedure was employed to select the observations used in estimating recharge. This procedure was based on previous work (Crosbie et al., 2018; Crosbie and Rachakonda, 2021) and involved the following steps: 1.Duplicate records were removed. 2.The mean value of chloride in groundwater was calculated for each location where multipleanalyses had been undertaken. 3.Any location where the excess water (rainfall minus actual evapotranspiration) was less than–25 mm/year was excluded. Negative excess water means the actual evapotranspiration isgreater than the rainfall, which indicates an additional source of water and chloride, forexample groundwater discharge, irrigation or flooding, etc. These violate the assumptions of theCMB method (see Section 3.8.1). 4.Any bore that was greater than 100 m deep was excluded. The ideal sampling point forestimating recharge would be immediately below the watertable where water has rechargedthe groundwater store recently. Monitoring bores are not constructed this way, so limiting thedepth is a compromise to ensure that water is sampled from outcrop areas and is as recent aspossible. 5.Chloride in groundwater concentrations of less than 2 mg/L were excluded, as thesemeasurements are probably not representative of groundwater. These analyses are closer tothe concentration of rainfall and probably indicate a mis-transcribed value entered in thedatabase, or a sample contaminated with rainfall through short circuiting of the bore casing. 6.Chloride in groundwater concentrations of greater than 2000 mg/L were excluded as these dataare probably not representative of groundwater in recharge areas. These are more likely to bedischarge areas, or downstream of discharge areas. 7. Bores that were located on alluvium as mapped in Geoscience Australia’s surface geology maps (Raymond et al., 2007) were excluded, as the alluvium is likely to have additional chloride present that is not directly from rainfall. This includes areas subject to flooding, irrigation or groundwater discharge from deeper or adjoining layers. 3.8.2 Upscaling point estimates of recharge The upscaling of the point estimates of recharge using regression kriging (Hengl et al., 2004) included three steps (detailed in subsequent sections): 1. developing a regression equation to predict recharge across the study area using covariates that can be mapped on a regular grid 2. kriging the residuals between the point estimates of recharge and the regression equation estimates of recharge to provide a surface of residuals 3. adding the residual surface to the regression surface, which provides a spatial estimate of recharge that is informed locally by point data, and away from point data is dependent upon a global relationship between the covariates and recharge. This process is repeated for 1000 stochastically generated replicates to quantify the uncertainty in the recharge estimates, as outlined in Figure 3-4. Regression equations Previous studies have shown that recharge is better approximated by a log-normal distribution rather than a normal distribution (Cook et al., 1989; Eriksson, 1985), and this extends to its relationship with rainfall (Petheram et al., 2002). Reviews of recharge in Australia and globally have shown that recharge estimates are dependent upon rainfall, soil type and vegetation (Crosbie et al., 2010; Kim and Jackson, 2012; Scanlon et al., 2006) and these have been successfully used as covariates to upscale modelled point estimates of recharge (Crosbie et al., 2013). Surface geology has also been used as a covariate in upscaling recharge estimates (Crosbie et al., 2015; Crosbie et al., 2018). Based on these previous studies, the covariates used are: • annual mean rainfall over the 30-year period between 1991 and 2020 (Jones et al., 2009) (Figure 3-8a) • mean clay content of the top 2 m of the soil profile (Grundy et al., 2015) (Figure 3-8b) • Normalised Difference Vegetation Index (NDVI) as a measure of vegetation density (BoM, 2019) (Figure 3-8c) • simplification of the surface geology (Raymond et al., 2012) (Figure 3-8d). Figure 3-8 Covariates used in upscaling (a) rainfall, (b) clay content of the soil, (c) Normalised Difference Vegetation Index (NDVI) and (d) simplified geology The regression equation used in the upscaling uses the covariates of log transformed rainfall (log10𝑃𝑃), clay content (C) and NDVI as a measure of vegetation density (V), and is: log10𝑅𝑅=𝛽𝛽geo+𝛽𝛽1.log10𝑃𝑃+𝛽𝛽2.𝐶𝐶+𝛽𝛽3.𝑉𝑉 (5) where βgeo, β1, β2 and β3, are fitting parameters fitted through least squares regression (which will be different for each of the 1000 replicates). The βgeo parameter is a set of three values representing the intercept value for each of the three geology classes (high, medium and low). Now consider a single replicate (that is a single choice for the chloride deposition and runoff) and the most successful covariates. To form the regression equation for this replicate, bootstrapping is used to reduce the effect of outliers and estimate the uncertainty in the regression equation (Efron and Tibshirani, 1994). That is, the same number of observations are randomly selected from the population of all observations, with replacement. This ensures that the same number of observations is used in the regression equation for each replicate but not every point will be For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au included in each replicate, and some points will be included more than once. This minimises the bias in the regression equation that can occur due to outliers in the dataset. Kriging residuals The difference in the recharge estimated through the regression equations and through the point- scale CMB are calculated for each replicate. The residual at each observation point is defined by: 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅=𝑙𝑙𝑙𝑙𝑙𝑙10(𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓)−𝑙𝑙𝑙𝑙𝑙𝑙10(𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝) (6) Here ‘fitted’ is the recharge (in millimetres per year) according to the regression equation, while ‘predicted’ is the recharge (in millimetres per year) according to the CMB method (Section 3.8.1). Hence if the residual equals 0.5, the regression equation overestimates the recharge by a factor of 100.5≈3. These residuals are fitted to a spherical semi-variogram using gstat (Pebesma, 2004) in R (R Core Team, 2016) and then kriged to a 0.01° (~1000 m) grid to create a residuals surface. The nugget, sill and the range parameters are fitted to the variogram separately for each replicate. Aggregating recharge rasters For each replicate, the residual raster was added to the recharge raster upscaled using the regression equations to create a regression-kriging recharge raster, at a 0.01° (~1000 m) resolution. Summary statistics were calculated from the 1000 replicates at the 5th, 50th and 95th percentiles, with the 50th percentile being assumed to be the ‘most likely’ estimate of recharge and the difference between the 5th and 95th percentiles a measure of the uncertainty. Constraining the uncertainty The CMB estimates of net recharge are constrained by the baseflow and the excess water estimates using a rejection sampling approach (Tarantola, 2005; Von Neumann, 1951). This is the same process used to constrain the CMB estimates of recharge in the Cambrian Limestone Aquifer (CLA) (Crosbie and Rachakonda, 2021) and relies on the baseflow being less than the recharge and the excess water being greater than the recharge. Three catchments were selected that are known to have high levels of baseflow and drain large parts of the study area (Drysdale et al., 2002; Yin Foo and Matthews, 2001). The selected gauges were the first gauges downstream of the known groundwater discharge zones (Figure 3-9a): Roper River at Elsey Homestead (G9030013); Gregory River at Riversleigh No. 2 (912105A); and Lawn Hill Creek at Lawn Hill No. 2 (912103A). The baseflow separation method used here is the digital recursive filter suggested by Lynne and Hollick (1979), although it has no physical basis it is the most commonly applied method of baseflow separation in Australia (Grayson et al., 1996). The baseflow separation was undertaken using the Basejumper program (Murphy et al., 2008) with a filter parameter of 0.925. It was assumed that the error in the baseflow estimates could be up to ±30% based on studies in other regions (Coxon et al., 2015; Petersen-Øverleir et al., 2009). The CMRSET algorithm for actual evapotranspiration (AET) (Guerschman et al., 2011) was used with MODIS data to create an 8-day time series at a 250 m resolution over the study area. These data were aggregated to a long-term mean for the period between 2001 and 2018. The long-term mean AET was subtracted from the long-term mean rainfall over the same period (Jones et al., 2009) to create a long-term mean excess water data layer. The long-term mean excess water was extracted for the three catchments used for the baseflow separation (Figure 3-9a) and for three selected internally draining catchments (BoM, 2012) (Figure 3-9b). Based on an evaluation of CMRSET water balances against runoff from stream gauges (King et al., 2011) and reviews of remotely sensed evapotranspiration (ET) (Glenn et al., 2011; Kalma et al., 2008), an error of up to ±30% has been assumed. Figure 3-9 (a) Catchments used for baseflow analysis for rejection sampling and (b) additional internally draining catchments used with excess water for rejection sampling GW = groundwater. Both the baseflow (BF) and excess water have an estimated uncertainty of ±30%. Assuming that the uncertainty is normally distributed, using the ‘six‐sigma rule’ gives the standard deviation of the baseflow prediction as 10% of the estimated value. This can then be used to estimate a probability distribution, which can be randomly sampled from: 𝐵𝐵𝐵𝐵𝑖𝑖=𝜇𝜇𝑖𝑖+𝑧𝑧.𝜎𝜎𝑖𝑖 (7) where z is the randomly selected standard normal deviate and μi and σi are the mean and standard deviation of the baseflow for the ‘ith’ catchment. In this way the same standard normal deviate is used to estimate the baseflow for each of the three catchments for each repetition of the rejection sampling algorithm. The uncertainty in the excess water is defined similarly: 𝐸𝐸𝐸𝐸𝑖𝑖=𝑃𝑃−(𝜇𝜇𝑖𝑖+𝑧𝑧.𝜎𝜎𝑖𝑖) (8) The mean recharge across the six catchments has been extracted from the 1000 upscaled replicates of the CMB. The rejection sampling algorithm was run 10,000 times with a randomly A map of different countries/regions Description automatically generated selected standard normal deviate for the baseflow distribution, a second randomly selected standard normal deviate for the excess water distribution and a randomly selected run number from the 1000 replicates of upscaled CMB recharge. A selection was retained in the posterior distribution if the following constraints were met: •the upscaled CMB estimates of recharge were greater than the baseflow estimates for each ofthe three catchments •the upscaled CMB estimates of recharge were less than the excess water estimates in each ofthe three additional internally draining catchments. 3.8.3 Extracting recharge values for zones of interest The 1000 m resolution rasters produced may be used to estimate recharge aggregated over any area. Mean recharge rates have been extracted for the major aquifers (Figure 3-2) and also the simplified surface geology (Figure 3-8d). 3.9 Identifying potential groundwater discharge areas using remote sensing Identifying areas of groundwater discharge has conventionally involved extensive observations in the field; this includes identifying the presence of phreatophytes, springs and seeps on the ground or increases in baseflow along river reaches. These approaches are labour intensive and impractical at the regional scale. As an alternative, remote sensing products offer an opportunity to identify and delineate the areas of potential groundwater discharge for further investigation in the field. 3.9.1 Digital Earth Australia The Digital Earth Australia (DEA) suite of products is derived from the Landsat Datacube produced by Geoscience Australia (Lewis et al., 2017). This dataset includes all Landsat imagery at a 25 m resolution dating back to 1987. There is a water bodies dataset (Krause et al., 2021) and a Water Observations from Space (WOfS) dataset (Mueller et al., 2016) that are both potentially useful. The WOfS dataset classifies each pixel in each Landsat image as being inundated with water or not. The dataset contains a summary layer that is the proportion of images from the Datacube that a pixel is inundated. In this way areas that are permanently inundated can be discriminated from areas that are infrequently inundated (and dryland areas). The DEA water bodies dataset is a polygon representation of contiguous areas that are inundated in more than 10% of images with a minimum area of five pixels (~0.3 ha). By identifying permanent water bodies there is the potential to narrow the search for groundwater discharge areas. The maximum from each pixel within a water body of the proportion of inundated images has been assigned to each polygon. In this way the permanence of a water body can be established. There is an issue of scale with this method, a large water body will be assigned a degree of permanence based on a single 25 × 25 m pixel. 60 | Characterising groundwater resources in the Southern Gulf catchments Any water bodies that can hold water throughout the dry season can be further investigated for the source of water. This could be surface water that replenishes a deep waterhole during the wet season and the depth of water is greater than the evaporation through the dry season leading to the permanent presence of water. Alternatively, it could be a continuous discharge of groundwater throughout the year providing a permanent presence of water. This process would also identify anthropogenic water storages such as instream or farm dams. 3.9.2 Excess water The CMRSET v2.2 dataset (Guerschman et al., 2022; McVicar et al., 2022) uses high-resolution / low-frequency satellite data (e.g. Landsat and Sentinel-2) blended with low-resolution / high- frequency satellite data (e.g. MODIS and VIIRS) to produce a monthly actual evapotranspiration product at a resolution of 30 m from 2002 onwards. By subtracting the annual mean actual evapotranspiration (AET) from the annual mean rainfall (P), excess water is calculated. Areas with positive values of excess water (i.e. P > AET) are exporting water elsewhere, either through runoff or through groundwater recharge. Areas with negative values of excess water (i.e. P < AET) are importing water, which can have a surface source such as floodplains and wetlands or a groundwater source such as springs and seeps. Areas with high negative excess water have been used previously to identify groundwater discharge areas (Crosbie and Rachakonda, 2021). These areas are indicative of having a water source throughout the year rather than only in the wet season for the majority of the landscape that is reliant upon rainfall. 3.9.3 Potential discharge areas over entire catchments In the Roper River Water Resource Assessment (Taylor et al., 2024) it was found that the permanence of water bodies using WOfS data and the excess water using CMRSET data were insufficient to identify groundwater discharge areas on their own. A method was developed using a threshold mean October AET to identify areas where dry-season AET was higher than expected and therefore potentially discharge groundwater. This threshold was found to be climate dependent with a higher threshold in the more tropical areas and a lower threshold in the more arid areas. If this threshold is defined as a ratio of AET / potential evapotranspiration (PET) and is scaled based on aridity index (P/PET), then the result is a threshold in the same form as the Budyko equation (Budyko, 1974): 𝐴𝐴𝐴𝐴𝐴𝐴 𝑃𝑃𝑃𝑃𝑃𝑃 =1+ 𝑃𝑃 𝑃𝑃𝑃𝑃𝑃𝑃 −􁉈1+􀵬 𝑃𝑃 𝑃𝑃𝑃𝑃𝑃𝑃 􀵰 𝑤𝑤 􁉉 1𝑤𝑤􀵗 (9) In this application the exponent w is not the same as used in the Budyko equation for estimating catchment AET and subsequent runoff. It is an empirical scaler and was determined through trial and error to be 1.65 based on known groundwater discharge areas in the Victoria, Roper and Southern Gulf catchments (Figure 3-10). Figure 3-10 Relationship used for determining the threshold October actual evapotranspiration (AET) for estimating potential discharge areas from aridity index PET = potential evapotranspiration; P = annual average rainfall. The areas identified by this threshold approach include all areas that have a high October evapotranspiration, irrespective of source. This includes floodplains and wetlands that can retain water through the dry season and actively managed areas such as dams and irrigation areas. Groundwater discharge areas tend to be a stable water source and do not exhibit high inter- annual variability. Wetlands and floodplains tend to have higher inter-annual variability due to their dependence upon retaining water from the previous wet season. This distinction is used to further aid in the identification of groundwater discharge areas using the coefficient of variation of the October evapotranspiration. A threshold of 0.5 has been used, with areas identified with high October evapotranspiration and high variability excluded as probably not being groundwater discharge areas. The CMRSET algorithm is over estimating evapotranspiration from bare soils, primarily dark or red soils. If it is assumed that areas of groundwater discharge are more likely to be vegetated, then false positives due to bare soils can be minimised by removing areas with high fractions of bare soils. This was done using the DEA Fractional Cover dataset (GA, 2021). This dataset identifies areas of green growing vegetation, brown dead or senescing vegetation and bare soil. Areas with a mean October bare soil fraction of greater than 15% have been excluded. To minimise isolated pixels that are more likely to be false positives, identified areas were collated as contiguous polygons and any that were less than 0.5 ha in size were excluded. There were still some false positives in the dataset, so each polygon was inspected in relation to geological contacts, its topographic position, and depth to watertable where these data exist. Each polygon was classified as either: • perennial groundwater discharge • seasonally varying • coastal • recharge feature • mis-identified For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au The ‘perennial groundwater discharge’ class most commonly represents springs and seeps. These can be on streamlines but are frequently not. They are usually associated with structural features such as fractures, joints and faults, or arising from change in geological units (one unit onlaps another), or at a break of slope. They are generally not in the alluvial areas low in the catchments. The ‘seasonally varying’ class is associated with the thin and storage-limited alluvium of the major rivers. They are conceptualised as discharging groundwater during the dry season from an alluvial store that is recharged via surface water in the wet season. The ‘coastal’ class is defined as polygons on estuarine or delta deposits or on the shoreline itself. These areas may have a groundwater discharge component as well as the tidal sea water component. ‘Recharge features’ hold water through the dry season but are above the watertable, these are frequently associated with sinkholes. The last class, ‘mis-identified?’, meet the criteria associated with the remote sensing products but do not have the geological attributes of the other classes and are not inferred to be discharge features. 4 Detailed, desktop and modelling investigations Section 4 of the report describes the methods used to conduct detailed desktop and modelling investigations of the Cambrian Limestone Aquifer (CLA) and Gilbert River Aquifer (GRA) including: (i) data related to the hydrogeological framework for the aquifers, (ii) data including groundwater levels and pressures, general chemistry and environmental tracer data to characterise groundwater flow processes, and (iii) data to be used as input into numerical modelling. 4.1 Hydrogeological framework 4.1.1 Geological modelling The compilation of pre-existing three-dimensional geological models, lithological and stratigraphic data and digital elevation models described in sections 3.1.2 and 3.3.1 formed the foundation of the development of local and regional three-dimensional geological models of the Georgina Basin and the Carpentaria Basin (a sub-basin of the Great Artesian Basin, GAB) using GoCAD/SKUA three- dimensional geological modelling software (Paradigm). In addition to existing surfaces, binding polygons that represent the extent of each hydrostratigraphic formation were developed to constrain the model development process. The models were developed using GoCAD/SKUA’s stratigraphic and structural modelling workflows described in detail by Raiber et al. (2019). In addition to the three-dimensional geological models of the Georgina Basin and geological Carpentaria Basin, simple three-dimensional representations of alluvial aquifers within the Gregory–Nicholson River alluvial systems were developed using lithological, stratigraphic and digital elevation data and World Image geotifs (sourced via Global Mapper geospatial software). The three-dimensional geological models were then integrated with other geospatial data types (e.g. watercourses, springs, potentiometric surfaces developed in the companion technical report on groundwater modelling (Knapton et al., 2024) and chloride mass balance (CMB) recharge estimates) to refine conceptual hydrogeological models and support the assessment of the suitability of different aquifers for groundwater development. 4.1.2 Hydrogeological cross-sections Hydrogeological cross-sections through the Carpentaria and Georgina basins were generated through integration of pre-existing formation boundaries (Section 3.1.2), stratigraphic logs and the three-dimensional geological models developed during the current study. A cross-section through the Carpentaria Basin was developed along an inferred flow path in the Gilbert River Formation based on potentiometric surfaces developed by Ransley et al. (2015). The potentiometric surface was also integrated into the three-dimensional geological modelling framework to determine potential changes of hydraulic conditions (e.g. sub-artesian versus artesian pressures) from the basin margin towards the deeper part of the basin. 4.1.3 Depth to top of major aquifers Depths to the tops of major aquifers in this study are based on Geoscience Australia data compilations, cross-sections and three-dimensional geological models. 4.2 Groundwater recharge and flow 4.2.1 Groundwater levels and bore yields Groundwater levels and bore yields have been extracted from the Department of Environment, Parks and Water Security (2000b) data files titled ‘bores_groundwater_level’ and ‘bore’ or ‘aquifer’ respectively for the NT. The Queensland groundwater-level and bore-yield data were extracted from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) within the data files titled ‘water levels’ and ‘aquifer’. 4.2.2 General chemistry Groundwater field parameters and chemistry were used to investigate recharge and mixing processes along flow paths in Section 6.2. These data were extracted from the ‘bore_water_quality’ data file from the Department of Environment, Parks and Water Security (2000b) for the NT. The Queensland data were extracted from the ‘water analysis’, ‘water_quality_field’ and ‘water quality variables’ data files from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023). 4.2.3 Environmental tracers Stable isotopes of water and radiocarbon were used to characterise recharge and groundwater origin and to confirm mixing processes (Section 6.2). These data were downloaded from the Geoscience Australia Data Portal (Schroder et al., 2020). 4.3 Groundwater salinity and hydrochemistry by stratigraphic formation A subset of the regional groundwater chemistry dataset was used for the Cambrian dolostone and limestone investigation in Section 6.3.1, and then for the Gilbert River Formation and Rolling Downs Group investigation in Section 6.3.2. Groundwater monitoring bores were selected both within and outside the Southern Gulf catchments to provide a better representation of regional groundwater chemistry due to the limited spatial coverage of available data within the catchments. 4.3.1 Cambrian dolostone and limestone The data presented in Section 6.3.1 represent a subset of the geochemical data, presented in Section 5.2, that is from the Cambrian dolostone and limestone stratigraphic class. This subset is a geographical subset exported from a model survey area called Undilla groundwater flow model in Section 4.4. Only water samples with a charge balance error (CBE) within ±10% were used to characterise the Cambrian dolostone and limestone aquifer. Due to the scarcity of data, all hydrochemistry time series data have been used to characterise hydrochemical patterns and processes within the Cambrian dolostone and limestone. However, for statistical calculation, only the most recent hydrochemistry results that meet data quality checks have been selected for each bore. Piper diagrams for the groundwater chemistry records were constructed using AcQA software (Rockware). 4.3.2 Gilbert River Formation and Rolling Downs Group As for the regional groundwater chemistry dataset (described in Section 3.6), only water samples with a CBE within ±15% were used. A relatively high CBE was chosen due to the sparsity of data in some parts of the Southern Gulf catchments. Furthermore, only the most recent hydrochemistry results were used where multiple analyses were available for a bore. Piper diagrams for the groundwater samples were constructed using AcQA software (Rockware). 4.4 Numerical flow modelling Unless stated otherwise, the following methods summary is based on a companion technical report on groundwater flow modelling by Knapton et al. (2024). This study along with a data review conducted by Knapton et al. (2024) undertook an extensive review of the hydrogeological data and information to confirm the conceptual model of the Cambrian Limestone Aquifer (CLA) in the eastern Georgina Basin. This conceptual model was used to develop an initial two-dimensional groundwater flow model for the CLA in the south-western part of the Southern Gulf catchments (the Undilla Sub-basin in the eastern part of the Georgina Basin). The purpose of the model was to assess the nature and scale of the groundwater resources of the Undilla Sub-basin, which provides baseflow to parts of Lawn Hill Creek and Gregory River. Initial groundwater balance estimates were derived for the entire groundwater model domain, which extends south-west of the catchments boundaries, and also for the portions of the Lawn Hill Creek and Gregory River subcatchments, and Nicholson Groundwater Management Area (NGMA), which coincide with the CLA inside the Southern Gulf catchments (Figure 4-1). Figure 4-1 Location of the Undilla Sub-basin groundwater flow model in relation to the Southern Gulf catchments and portions of the model that coincide with the Lawn Hill Creek and Gregory subcatchments and Nicholson Groundwater Management Area (NGMA) Lower left inset indicates the south-western part of the catchments where the mapped data are shown, as well as its location within the eastern part of the Georgina Basin (light purple). GW = groundwater. 4.4.1 Model description The groundwater flow model of the unconfined and confined areas of the CLA within the Undilla Sub-basin is referred to as the Undilla1 groundwater flow model. The Undilla1 groundwater model covers an area of 50,900 km2 and its extents are presented in Figure 4-1. The northern and eastern boundaries are coincident with the margin of the Georgina Basin, the northwestern boundary is coincident with the Alexandria–Wonarah Basement High and the western and southern boundaries are coincident with the mapped occurrence of the Camooweal Dolostone (represented by the Cambrian dolostone in Figure 2-11). The Undilla1 groundwater flow model of the CLA in the Undilla Sub-basin is a two-dimensional, single layer finite element numerical model, developed using the FEFLOW simulation code (Diersch, 2008). The CLA groundwater system is conceptually characterised as an equivalent A map of the united states Description automatically generated porous medium. This simplification allows for the development of a more manageable and computationally efficient model while still capturing the essential characteristics of the groundwater system using calibrated regional aquifer parameters to reproduce the observed groundwater levels and discharge to the rivers. This assumption means that the actual flow paths cannot be modelled and that there is no intention for this model to be used for contaminant transport purposes. The finite elements are roughly equidimensional (mean = 28 km2, significant difference = 12 km2) with some refinement around the Lawn Hill Creek and Gregory River (approximately 1 to 2 km2). The finite element mesh geometry showing pilot point locations and specified head boundary conditions along Lawn Hill Creek, Gregory River and the throughflow boundary to the south is presented in Figure 4-2. Figure 4-2 Finite element mesh geometry showing pilot point locations and identifiers and specified head boundary conditions along Lawn Hill Creek and the Gregory River, and the throughflow boundary to the south For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Recharge is applied to the top slice of the groundwater model based on long-term annual mean recharge estimates derived from the CMB method to scale the rainfall record. Rainfall data were sourced from the SILO Data Drill for a site located at latitude –19.50°S and longitude 138.00°E, extracted on 22 March 2024. The recharge estimate is empirical and does not account for processes such as preferential or bypass flow. The transmissivity parameter field was applied to the elements in the model via pilot points and the interpolation code PLPROC (Doherty, 2024). Pilot points were placed throughout the model domain at a 25 km x 25 km spacing to allow for a flexible spatial parameterisation; they play a pivotal role in the calibration process. The pilot point locations and their identifiers are presented in Figure 4-2. For this initial assessment, the storage coefficient was applied as a single value to the entire model domain. The groundwater model includes boundary conditions that define the interaction between the rivers and the groundwater system. Discharge from the rivers is implemented using specified head boundary conditions. The locations of the specified head boundary conditions are presented in Figure 4-2. Discrete springs are not included in the model as pathways are too poorly understood and at a scale too small to be adequately represented. Extraction for stock, domestic and horticultural use is not included. 4.4.2 Model calibration The calibration process utilised a combination of pilot points (Doherty, 2003) and PEST_HP (Doherty, 2024). PEST_HP, an automated nonlinear parameter estimation code, iteratively calibrated the model by adjusting the pilot point parameters and recharge parameter to minimise the objective function (i.e. discrepancies between observed and simulated data). The observed data used to define the objective function included available historical groundwater levels (474 head observations) in the CLA and discharge measurements: 638 for the Gregory River (912101A) and 242 for Lawn Hill Creek (912103A). The discharge measurements were weighted to emphasise the dry-season values, which are assumed to represent groundwater discharge. Groundwater levels at each bore consist of a single water level recorded at the time of construction. This has been assumed to correspond to the end of the dry season for the year 2019. This assumption is likely to introduce a significant bias where bores were installed during periods not representative of recent conditions. 4.4.3 Simulation runs The model was run monthly for the 109-year climate sequence from 1910 to 2019. Model reporting was conducted over the proposed water management zones of the Water Resource (Gulf) Plan 2007. Water balances have been reported for the Lawn Hill subcatchment, Gregory subcatchment, the NGMA identified in the Water Resource (Gulf) Plan 2007 and the entire model domain (Figure 4-3). The outputs of the modelling have been reported as: • water levels for eight groundwater reporting sites (Figure 4-3) • groundwater contours over the model domain • groundwater discharge as a combination of discharge from springs and lateral outflow where the streams are incised into the CLA along the Lawn Hill Creek (912103A) and the Gregory River (912101A) • water balances, documented for four areas within the model domain: the entire model domain, Lawn Hill catchment, the Gregory catchment and the NGMA. Figure 4-3 Location of groundwater head and flow reporting sites and the Nicholson Groundwater Management Area used for water balance reporting For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Part III Results Nicholson River Photo: CSIRO 5 Regional desktop and modelling assessment of the Southern Gulf catchments Section 5 of the report presents the results of the regional desktop and modelling assessment of the Southern Gulf catchments. 5.1 Aquifer data Basic summary statistics of groundwater bore construction information (i.e., total depth and screen length) are provided in Table 5-1. Bores assigned to the Cambrian dolostone and limestone stratigraphic class range in depth from 24 to 537 m, with a median of 101 m. Groundwater bores assigned to the Rolling Downs Group have a median bore depth of 135 m, whereas the underlying Gilbert River Formation has a median bore depth of 309 m within the extended catchment of the Southern Gulf rivers (with a minimum of 28 m and a maximum of 747 m). The summary unit ‘igneous rocks’, which is composed of many different rock types, has a median total bore depth of 75 m, but a wide range of values from 18 to 786 m depth. The ‘alluvium and colluvium’ stratigraphic unit has a median bore depth of only 23 m reflecting that the alluvium and colluvium are relatively shallow (with intersected alluvial thicknesses in groundwater bores within the coastal parts of the Southern Gulf catchments of up to 25 m described by Buchanan et al. (2020). Screen length of bores was also assessed. These are of particular importance when assigning bores to aquifers and when planning field sampling campaigns. For example, where groundwater bores have very long screens, these often extend over multiple stratigraphic units. When using hydrochemistry, yields or standing water levels (SWLs) to infer differences between different aquifers, then short screens are preferable as they are more likely reflective of a single aquifer. Within the extended Southern Gulf catchments, median screen length of different simplified stratigraphic units ranges from 2 m (for the alluvium and colluvium) to 30 m (Gilbert River Formation). However, the range of the length is highly variable. For example, for the Gilbert River Formation, the largest screen length recorded (assumed from the top of the shallowest screen to the bottom of the deepest screen) is 480 m. This indicates that this bore may be screened not only within the Gilbert River Formation but also within other shallower Great Artesian Basin (GAB) formations. 5.2 Groundwater levels Median values of groundwater levels or pressures were calculated for each of the major stratigraphic classes included in this study (Table 5-1and Figure 5-1). Average, median, minimum and maximum standing water level were calculated for each formation to provide insights into the degree of spatial variability within different stratigraphic classes. The median of the standing water level (or groundwater pressure where aquifers are confined) is negative for most formations within the extended Southern Gulf catchments. This indicates that the groundwater level or pressure is below the ground surface. For the Cambrian dolostone and limestone, for example, the median value of the standing water level is –61 m (corresponding to 61 mBGS), but ranges from –156 m (minimum) to –5 m (maximum) (Table 5-1). As expected, for the alluvium and colluvium, the median standing water level is much closer to the surface (12 mBGS). An exception to this is the Gilbert River Aquifer, hosted within the Gilbert River Formation, where the median value is positive (3 m, with a minimum of –200 m and a maximum of 35 m), indicating artesian conditions in some parts of the aquifer. This was also highlighted by previous studies (e.g. Buchanan et al., 2020), which suggested that flowing artesian bores such as the ‘Burketown bore’ (with Queensland bore registration number RN330) are known to exist in the area. Figure 5-1 Depth to standing water level within the extended Southern Gulf catchments Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014); Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) A map of the united states Description automatically generated 5.3 Groundwater salinity and hydrochemistry 5.3.1 Groundwater salinity Summary statistics of groundwater salinity data (minimum, maximum, mean and median salinity (expressed by the total dissolved solid (TDS) concentrations in milligrams per litre) are shown in Table 5-1 for each of the major simplified stratigraphic units. The formation with the widest range of salinities is the Proterozoic undifferentiated sedimentary rocks, which has both the lowest (24 mg/L TDS) and highest groundwater salinities (~27,000 mg/L TDS) recorded within the study area (with a median of 525 mg/L) (Figure 5-2). Table 5-1 Overview statistics on bore construction and basic hydrogeological and hydrochemical characteristics of aquifers For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au N/A = not applicable Figure 5-2 Spatial distribution of total dissolved solids (TDS) concentrations for major aquifers within the extended Southern Gulf catchments The salinity of groundwater samples within the extended Southern Gulf catchments ranges from extremely fresh (24 mg/L TDS) to saline (with the highest recorded TDS of 27,000 mg/L) (Table 5-1). Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014); Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) A map of the united states Description automatically generated 5.3.2 Groundwater chemistry Multivariate statistics – major stratigraphic classes This section describes the results of multivariate statistical analyses of groundwater chemistry based on records sourced from the Queensland groundwater database and NT groundwater database (Section 3.6). Groundwater chemistry can provide useful insights into a range of important groundwater processes, including: • groundwater recharge • hydrochemical evolution within aquifers (e.g. influence of ion exchange or methanogenesis) • connectivity between different aquifers • surface water – groundwater interactions. Multivariate statistical analyses of groundwater chemistry records (hierarchical cluster analysis; HCA) were conducted to make inferences about the likely processes that control spatial and temporal variability of water chemistry. The results of the HCA are presented in Table 5-2 and Figure 5-3, showing that there are five major clusters with distinct median values for different parameters. A chi-square test was used to test for independence between aquifers and clusters and assess whether the cluster attribution of hydrochemical records is related to aquifer (or surface water) membership. Since the P-value is less than 0.05, the hypothesis that the aquifer and cluster classifications are independent can be rejected at the 95% confidence level. The relationship between cluster and aquifers is highlighted in a cross-tabulation (Figure 5-3), which shows to which cluster different aquifers were assigned. Table 5-2 Median values of the variables considered in the cluster analysis for each sample group For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au TDS = total dissolved solids, Na = sodium, Ca = calcium, Mg = magnesium, K = potassium, Cl = chloride, SO4 = sulfate, F = fluoride, HCO3 = bicarbonate Median values for different parameters were used to determine the characteristics of the different water chemistry clusters. Figure 5-3 Cluster membership of aquifers in the Southern Gulf catchments. Total dissolved solids (TDS) corresponds to the median total dissolved solid concentrations of each cluster. The numbers correspond to the number of hydrochemical records assigned to each aquifer for major stratigraphic units A diagram of a rectangular object Description automatically generated with medium confidence The median concentrations of the identified hydrochemical clusters are also shown on Piper and Schoeller plots (Figure 5-4) to identify the major differences between clusters and the spatial distribution of hydrochemical clusters is shown in Figure 5-5. Figure 5-4 Median ion values of the hierarchical cluster analysis presented as (a) Piper plot and (b) Schoeller plot For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 5-5 Spatial distribution of cluster membership within the extended Southern Gulf area A map of the united states Description automatically generated The major characteristics of the different clusters are: •Cluster 1: this group contains a large number of groundwater chemistry records from theCambrian dolostone and limestone and a smaller number of samples from other stratigraphicformations. A large number of samples with unknown stratigraphic formation at the screenedinterval (e.g. due to a lack of lithological and stratigraphic data) and some alluvium andcolluvium and Cenozoic fluvial samples are also assigned to this cluster. With a median TDS of1254 mg/L, these groundwaters are slightly brackish. They are Na–Cl type waters but also haverelatively high concentrations of most other major ions (e.g. Ca, Mg, K, HCO3 and SO4) and amedian of 1.1 mg/L of fluoride. •Cluster 2: water chemistry records in this group are mostly from the Cambrian dolostone andlimestone, the Proterozoic undifferentiated sedimentary rocks and some records from theigneous rocks. Only very few records from the Rolling Downs Group and Gilbert River Formationwere assigned to this cluster. Water chemistry records assigned to this cluster are generallyfresh (with a TDS of 484 mg/L) Mg–HCO3 type waters, which have very high HCO3/Cl ratios and adominance of calcium and magnesium over sodium. •Cluster 3: the majority of water chemistry records in this group are sourced from the GilbertRiver Formation (n = 117), and the majority of Gilbert River Formation groundwaters areassigned to this cluster. These samples are slightly brackish (median TDS of 1098 mg/L) and areNa–Cl type waters; they also have high concentrations of HCO3, very low calcium andmagnesium concentrations and high concentrations of fluoride (median of 2.95 mg/L). Highfluoride concentrations in other sedimentary basins are commonly considered a potentialindicator of interaction with coal or organic matter within the sedimentary sequences. Thepresence of methane in some groundwaters of the Gilbert River Formation in this region wasalso highlighted by Buchanan et al. (2020). Overall, groundwaters assigned to this cluster arehighly evolved, and are potentially groundwaters with very long residence times. However, ashighlighted by Raiber et al. (2022), only very limited environmental tracer data exist in this partof the GAB. •Cluster 4: water samples assigned to this group have the highest median salinity within theSouthern Gulf catchments (median TDS of 3841 mg/L). They are mostly sourced from theCambrian dolostone and limestone and samples with an unknown aquifer outside theboundaries of the Southern Gulf catchments within the extended assessment area (Figure 5-5). Samples within this cluster have a low HCO3/Cl ratio, high concentrations of all cations and veryhigh SO4 concentrations. •Cluster 5: this group almost exclusively contains groundwater samples from the Proterozoicundifferentiated sedimentary rocks (60 out of 67 samples). Cluster 5 groundwaters are markedby extremely low salinities (median TDS of 35 mg/L). These groundwaters are classified as Na–Cltype waters with very low concentrations of all ions including fluoride. These samples are mostlycollected from groundwater bores in the northern part of the extended Southern Gulfcatchments and where the freshness of the groundwaters suggest that rapid recharge likelylimits the potential for evapotranspiration to occur. Six samples assigned to ‘Unknown’ aquifermay also likely be sourced from the Proterozoic undifferentiated sedimentary rocks. However, the cross-tabulation (Figure 5-3) shows that many samples from the Proterozoic undifferentiated sedimentary rocks are also assigned to other clusters, suggesting significant variability within these Proterozoic strata. 5.4 Bore yield Summary statistics of bore-yield data (minimum, maximum, mean and median salinity (expressed in litres per second)) are shown in Table 5-1 for each major simplified stratigraphic unit. Medians for most formations are very similar. However, the range of values (from minimum to maximum) differs significantly. For example, bore yields for the Cambrian dolostone and limestone range from 0.1 to 20 L/second with a median of 2.3 L/second. For alluvial aquifers, yields ranged from 0.7 to 3 L/second, with a median of 2 L/second. For the Cretaceous rocks of the Carpentaria Basin, estimated bore yields are mostly low (less than 5 L/second) (Figure 5-6), but values higher than 10 L/second were recorded for some bores in the Gilbert River Aquifer. Some bore yields were also recorded for a bore screened near Settlement Creek in the northern part of the Southern Gulf catchments. Estimates of yield were relatively low in this area (mostly less than 5 L/second). Figure5-6Spatial distribution of bore yields in major aquifers within the extended Southern Gulf catchments Data sources:NTG Data Portal (Department of Environment Parks and Water Security, 2014); Queensland groundwater database (Department ofRegional Development, Manufacturing and Water, 2023) 84|Characterising groundwater resources inthe Southern Gulfcatchments 5.5 Recharge estimation This section includes recharge estimates using the chloride mass balance (CMB) method; it follows the same format reported in Section 3.8. The results are presented at the point scale, then upscaled to the entire study area, constrained using observations of baseflow and excess water with final recharge estimates derived for various areas of interest. 5.5.1 Point-scale chloride mass balance The database of chloride in groundwater measurements contains 2319 locations with at least one measurement (Figure 5-7). The chloride in groundwater ranged from 2 to 25,000 mg/L with a median of 122 mg/L. At these points, the estimated mean chloride deposition in rainfall ranged from 54 kilograms per hectare per year along the coast to 2.0 kilograms per hectare per year further inland based on the Australian-wide rainfall chloride deposition map (Wilkins et al., 2022). This Australian-wide map showed that there is a lack of rainfall chloride observation points within the broader Southern Gulf catchments. Of these 2319 points, 1603 were retained after being assessed against the criteria in Section 3.8.1. The median of the 1000 replicates of point recharge at these 1603 points is shown in Figure 5-7. It ranges between 0.1 to 679 mm/year, with a mean of 45 mm/year and a median of 4 mm/year, demonstrating the skewed distribution of recharge. Figure 5-7 The chloride (Cl) in groundwater (GW) observations within the study area and the median of the point- scale estimates of recharge derived from them A map of the united states Description automatically generated 5.5.2 Upscaling point estimates of recharge Regression equations The median of the 1000 replicates of point recharge is used in an exploration of the covariates chosen for upscaling (Figure 5-8 and Figure 5-9). There is a positive correlation between rainfall and recharge (Figure 5-8a): as rainfall increases so does recharge. Clay content of the top 2 m of the soil profile shows a negative correlation (Figure 5-8b), with recharge decreasing as clay content increases. The relationship with the Normalised Difference Vegetation Index (NDVI) is not as expected where a higher vegetation density should result in lower recharge. The observed positive correlation (Figure 5-8c) is due to the relationship between rainfall and NDVI. As higher vegetation density is expected with higher rainfall, the relationship with rainfall needs to be accounted for before the negative correlation with recharge can be seen. The three geology classes show a similar slope of regression line (Figure 5-9) but differ in their intercepts. The high recharge class has the largest intercept, and the low recharge class has the smallest intercept. Figure 5-8 Point-scale relationships between recharge and (a) rainfall, (b) clay content of the soil and (c) Normalised Difference Vegetation Index (NDVI) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 5-9 Point-scale relationships between rainfall and recharge by geology class The combined results of the covariates in predicting recharge were analysed through multiple linear regressions. The parameter results are shown as boxplots in Figure 5-10. The value of βgeo for the high recharge class is greater (less negative) than the values for the medium and low recharge geology classes indicating that the high recharge class does indeed have higher recharge than the other classes, all other variables being equal. The coefficients for the clay content and NDVI are negative as they should be, as increases in clay content or NDVI should lead to decreased recharge. The results of the regression equation can be measured by the R2 as a measure of the goodness-of-fit of the models. The R2 had a mean of 0.59 and a range of 0.45 to 0.72 for the 1000 replicates. This means that the regression equation is only able to explain (on average) 59% of the variance in the recharge estimates. This is probably not a reflection of a poor model but a noisy dataset, and is similar to other predictions of recharge using regression analysis (Crosbie et al., 2018; Crosbie and Rachakonda, 2021). For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 5-10 Coefficients used in the regression equations for upscaling the 1000 replicates (a–f), (g) the R2 for each of the 1000 replicates The line in the centre of the box is the median, the box represents the interquartile range (the 25th to 75th percentiles), the whiskers represent the 10th and 90th percentiles and the dots are any data points outside the 10th and 90th percentiles. NDVI = Normalised Difference Vegetation Index. Kriging residuals The point estimates of recharge can be upscaled using the fitted regression equations and the spatially mapped covariates. The mean of the 1000 replicates is shown in Figure 5-11a. The spatial trends in the recharge follow the rainfall trends, with the highest recharge in the areas with the highest rainfall and vice versa. The residuals of the fitted recharge derived from the regression equation were kriged to create a surface. The mean of the 1000 replicates is shown in Figure 5-11b. This surface has a mean of zero and has a mean absolute residual of the upscaled surface of 0.21. In areas without data points, the residual surface tends to zero; this is expected as the recharge estimates were derived from the regression equations. In areas with data points, the regression estimates will be ‘corrected’ towards the point values of recharge. The variogram used has a nugget fixed at 0.1, which prevented over-fitting; this can be seen in the lack of ‘bullseyes’ around each data point. The range of the variograms is approximately 50 km, this is the distance that each data point can influence. If some structure can be found in the patterns of the residual surface, then this could be brought into the regression equations to reduce the residuals and increase the fit of the regression equations, but the cause of the patterns could not be identified in this study. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 5-11 (a) The median value of the 1000 replicates of upscaled recharge using the regression equation and (b) the median of the residuals kriged to a regular grid also showing the points used Aggregating recharge rasters When the residuals were added back to the regression estimates of recharge, the final set of regression-kriging upscaled recharge estimates was created. These are displayed as the 5th, 50th and 95th percentiles of the 1000 replicates in Figure 5-12. Over the entire region being investigated, the surfaces representing the 50th, 5th and 95th percentile recharge have mean values of 25 mm/year, 13 mm/year and 47 mm/year, respectively. A map of different countries/regions Description automatically generated Figure 5-12 The 5th, 50th and 95th percentiles of upscaled recharge from the 1000 replicates using regression kriging The 50th percentile upscaled recharge estimates on a 0.01° (~1000 m) grid are compared to the median point estimates of recharge in Figure 5-13. This shows that the mean R2 for the 1000 replicates has increased from 0.59 for the regression equation to 0.87 for the regression kriging, demonstrating a much better fit to the point estimates of recharge from the CMB method. Figure 5-13 (a) Scatterplot of the 50th percentile of the recharge calculated at a point scale versus the 50th percentile of the regression-kriging upscaled recharge estimate (red line is 1:1 for reference), and (b) the R2 of all 1000 replicates for the point estimates of recharge versus the upscaled estimates of recharge as a boxplot A map of different countries/regions Description automatically generated For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Constraining the uncertainty The 1000 replicates were constrained using the baseflow at the low end and the excess water at the high end in an attempt at reducing the uncertainty in the recharge estimates. Of the 10,000 random selections from the 1000 replicates, 5282 passed the rejection sampling criteria. This process only saw marginal improvement to the uncertainty with more samples being rejected at the low end (by the baseflow observations) than at the high end (by the excess water observations). The constrained 5th, 50th and 95th percentile estimates of recharge are displayed in Figure 5-14. Over the entire region being investigated, the surface representing the 50th, 5th and 95th percentile recharge have a mean of 27 mm/year, 15 mm/year and 49 mm/year, respectively. Figure 5-14 The 5th, 50th and 95th percentiles of constrained recharge for the modelled area The constrained recharge clipped to the Southern Gulf catchments is shown in Figure 5-15. The surface representing the 50th percentile recharge has a range between 1.5 and 574 mm/year with a mean of 23 mm/year. The surface representing the 5th percentile recharge has a range between 0.7 and 379 mm/year with a mean of 12 mm/year. and the surface representing the 95th percentile recharge has a range between 2.8 and 846 mm/year with a mean of 45 mm/year. At the catchment scale, Mornington Island has the highest recharge with a mean of 80 mm/year for the 50th percentile while the catchment of the Leichhardt River has the lowest with 7.5 mm/year. Figure 5-15 The 5th, 50th and 95th percentiles of constrained recharge for the Southern Gulf catchments 5.5.3 Extracting recharge values for zones of interest The recharge rates were extracted for the major hydrogeological units for the extent of the Southern Gulf catchments (Table 5-3). The recharge to the Thorntonia Limestone and Camooweal Dolostone is comparatively small with 50th percentile values of 10 and 11 mm/year, respectively; this is due to them being located in the lower rainfall parts of the study area. The highest 50th percentile recharge of 48 mm/year was in association with the Proterozoic carbonates. A map of different countries/regions Description automatically generated Table 5-3 Mean recharge rates over each of the major aquifers. The 50th percentile is outside the brackets and the 5th and 95th percentiles give a range for the uncertainty within the brackets For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au The mean recharge rates aggregated to the simplified surface geology classes is shown in Table 5-4. The highest recharge was seen in the volcanics class, which is relativity small and in thehighest rainfall part of the catchments. Table 5-4 Mean recharge rates over the simplified surface geology classes. The 50th percentile is outside the brackets and the 5th and 95th percentiles give a range for the uncertainty within the brackets For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 5.6 Identifying potential groundwater discharge areas using remote sensing 5.6.1 Digital Earth Australia The Digital Earth Australia (DEA) water bodies dataset (Krause et al., 2021) identifies 5286 water bodies that are greater than 0.3 ha in size and contain water more than 10% of the time (Figure 5-16). Only 480 of these water bodies contain water more than 75% of the time and can beconsidered (semi-) permanent. The majority of these occur along the major rivers; however, someare located in the coastal zone. This dataset has not identified springs of interest around Lawn Hill and to the west of the Settlement Creek catchment. This may be due to a scale issue as some springs might be smaller than the resolution of the data (25 × 25 m with a minimum of five pixels). Figure 5-16 Water bodies in the Southern Gulf catchments identified from Digital Earth Australia and the proportion of time that water bodies are inundated from Water Observations from Space The size of the water bodies has been exaggerated so that they can be seen at the catchment scale. 5.6.2 Excess water The map of excess water across the Southern Gulf catchments (Figure 5-17) shows which areas are exporting water (green) and which are importing water (purple). Areas that are importing water are potentially groundwater discharge areas (or floodplains, irrigation areas, etc.). There are large areas in the Leichhardt catchment that appear to have much higher actual evapotranspiration (AET) than rainfall. These areas are not in topographic lows and so appear to be overestimated AET rather than potential groundwater discharge areas. The other large area that can be seen with a high negative excess water is on the coast of the catchments of the Nicholson, Leichhardt and Morning Inlet rivers. These areas are potentially periodically inundated by sea water. The springs mapped in the Lawn Hill area and in the west of the Settlement Creek catchment have been identified as having AET greater than rainfall but cannot be seen at the catchment scale. Figure 5-17 Excess water across the Southern Gulf catchments P-AET = potential actual evapotranspiration. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 5.6.3 Potential discharge areas over the whole of the catchments There were 88,300 polygons identified in the Southern Gulf catchments as potential groundwater discharge features using the mean October AET. Of these, 15,367 were larger than 0.5 ha and were considered further. It is these larger potential discharge areas that are important at this regional scale when considering the water resource assessment. Smaller discharge features can potentially be identified down to 900 m2 pixel size, but these would be very localised in scale and difficult to consider at the regional scale. Not all of these 15,367 polygons are groundwater discharge features. Any permanent water bodies that are above the watertable and act as recharge features will also have been identified. Without an adequate depth to watertable map of the catchments these features cannot be algorithmically excluded and need to be manually excluded. Each polygon was inspected individually and assigned to one of five categories (Figure 5-18, Table 5-5): • The ‘perennial groundwater discharge’ category included 993 polygons for a total area of 1701 ha. These features are predominantly springs and soaks in the mid to upper reaches of catchments and are related to geological contacts. • The ‘seasonally varying’ category included 2624 polygons for a total area of 16,043 ha. These features are mostly associated with the alluvium in major rivers. They are mostly conceptualised as being recharged by surface water during the wet season and then discharging this water during the dry season through evapotranspiration. In the case of the Gregory and O’Shannassy rivers and Lawn Hill Creek, these streams are draining regional groundwater and are receiving groundwater throughout the year. • There were no polygons identified in the ‘recharge’ category. • The ‘coastal’ category included 2065 polygons covering a total area of 23,771 ha. These are in the estuarine part of the major rivers and along the coast of the Gulf of Carpentaria. These areas may have a component of groundwater discharge along with the evapotranspiration of sea water. • The ‘mis-identified?’ category included 9667 polygons covering a total of 19,008 ha. These features seem to have anomalously high October evapotranspiration without a geological explanation. Table 5-5 Summary of areas identified as potential groundwater discharge areas For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 5-18 Areas of potential groundwater discharge across the Southern Gulf catchments The size of the polygons has been greatly exaggerated to allow them to be seen at the catchment scale. Data sources: Spring databases of the NT Government (Department of Environment and Natural Resources (NT), 2019) and Queensland Government (Department of Environment, Science and Innovation, 2021) A map of the united states Description automatically generated The Springs databases of the NT Government (Department of Environment and Natural Resources (NT), 2019) and Queensland Government (Department of Environment, Science and Innovation, 2021) are the best comparison datasets to assess the performance of this method. These databases have been compiled over decades of fieldwork where springs have been identified in the field. They do not purport to be complete or to capture all the major springs. There are 60 springs identified within the government databases. The median distance between the springs in the databases and those identified here is 172 m. Considering the accuracy of the mapping in the database (some spring locations are noted as approximate, others as being hundreds of metres upstream), it can probably be concluded that most of the springs were correctly identified. Some non-identification of springs could be scale dependent, a threshold of 0.5 ha was used here so small seeps and soaks will not be identified. A better test for springs identified here would be field verification of the more than 900 springs identified that are not in the databases. 5.6.4 Summary of potential discharge areas from remote sensing The mapping of water bodies from DEA and assessing their permanence using Water Observations from Space (WOfS) is useful for identifying water bodies that are larger than 0.3 ha. The pixel size is 25 m from Landsat and therefore minor pools and waterholes within streams are often too small to be identified using this platform. Excess water calculated using CMRSET data identified many areas that had evapotranspiration that were much greater than rainfall but many of these were not groundwater discharge areas and are probably areas of overestimated AET. The use of October evapotranspiration from CMRSET provided a much more useful guide to the location of groundwater discharge areas. Areas that are still using a lot of water at the end of the dry season have another source other than rainfall, and in a lot of cases this will be groundwater. The method used here identified 993 potential springs across the Southern Gulf catchments compared to the 60 springs that have been documented from field studies in databases held by the NT and Queensland governments. This method successfully identified most of the springs in the database. There are expected to be differences due to the scale of investigation. The next step is field verification of the springs that are not in the government databases. 6 Detailed desktop and modelling investigations Section 6 of the report presents the results of the detailed desktop and modelling investigations of the Cambrian Limestone Aquifer (CLA) and Gilbert River Aquifer (GRA). 6.1 Hydrogeological framework 6.1.1 Cambrian dolostone and limestone As described in Section 2.4.2, the fractured and karstic carbonate rocks of the Georgina Basin which hosts the Cambrian Limestone Aquifer (CLA) offer potential to support regional development based on existing hydrogeological data but remain poorly characterised. New drilling conducted as part of South Nicholson National Drilling Initiative (Geoscience Australia, 2024) and newly developed three-dimensional geological model surfaces (Rollet et al., 2024) provide new insights into the composition and geometry of the Georgina Basin carbonate rocks hosting the CLA. Based on these regionally extensive three-dimensional geological model surfaces, simple three-dimensional geological models of the Georgina Basin and isopach (thickness) maps were developed to provide a regional understanding of the Cambrian dolostone and limestone that hosts the CLA within the Southern Gulf catchments. The basement topography underneath the Georgina Basin and the underlying South Nicholson Basin is complex, with multiple sub-basins separated by basement highs, including the Alexandria–Wonarah High (Figure 6-2a). The largest thicknesses of Georgina Basin strata within the Southern Gulf catchments are approximately 1000 m (Figure 6-1, Figure 6-2 and Figure 6-3). Figure 6-1 Thickness (isopach) map of Georgina Basin Data sources: Gallant et al. (2011), Carr et al. (2016), Vizy and Rollet (2024) A map of the north and south america Description automatically generated Figure 6-2 Three-dimensional geological model (30× vertically exaggerated) of Georgina Basin showing (a) basement topography, groundwater divides and inferred flow direction; (b) Georgina Basin strata, springs and watercourses within Southern Gulf catchments; and (c) thickness of Georgina Basin Data sources: Gallant et al. (2011), Carr et al. (2016), Department of Environment, Science and Innovation (2021), Vizy and Rollet (2024) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au At Carrara 1 stratigraphic well (see Figure 6-2 for location), recently drilled as part of the South Nicholson National Drilling initiative (Carr et al., 2019), the total thickness of the Georgina Basin is approximately 630 m. Stratigraphic logs and high-resolution mineralogical (X-ray diffraction, XRD) analysis (Figure 6-4) of Carrara 1 core samples (Figure 6-1) shows that there is considerable lithological variability within the Georgina Basin strata at this location. The shallowest formation, the Camooweal Dolostone, is approximately 250 m thick. It is almost exclusively composed of dolomite with only minor quartz (Figure 6-4). Underlying the Camooweal Dolostone at Carrara 1 is the Currant Bush Limestone. The Currant Bush Limestone is described as a bedded, partially dolomitised limestone and dolomitic limestone, with interbeds of ooid grainstone, shale, siltstone, marl and chert (Dixon-Jain, 2024). At Carrara 1, the Currant Bush Limestone is approximately 280 m thick. It unconformably overlies the Thorntonia Limestone, which outcrops at the surface in the eastern part of the Georgina Basin within the Southern Gulf catchments but has not been intersected at Carrara 1. The Thorntonia Limestone is composed of marine bioclastic carbonates, including limestone, partially dolomitised limestone, dolostone, pyritic-carbonaceous dolostone, marl and mudstone, with minor nodular chert and phosphorite (Dixon-Jain, 2024). The mineralogical analysis of the Currant Bush Limestone at Carrara 1 confirms that the composition of this formation is more variable than the Camooweal Dolostone, with significant proportions of calcite, K-feldspar, quartz and mica. Figure 6-3 Cross-section through the Georgina Basin. For orientation of cross-section, see Figure 6-1 Data sources: Gallant et al. (2011), Carr et al. (2016), Vizy and Rollet (2024) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-4 Mineralogical composition (X-ray diffraction, XRD) of Georgina Basin and underlying strata at stratigraphic well Carrara 1 (based on data from Owen et al., 2023) 6.1.2 Gilbert River Formation and Rolling Downs Group The Gilbert River Formation which hosts the Gilbert Rivber Aquifer (GRA) is the major GAB aquifer within the Southern Gulf catchments. Based on new three-dimensional geological models developed by Geoscience Australia (Section 3.1.2), the thickness of the Gilbert River Formation is typically less than 100 m in the Southern Gulf catchments but the thickness increases towards the north-west and offshore from the Southern Gulf catchments (Figure 6-8). The visual inspection of cores at the QLD core library provided first insights into the mineralogy of different formations: • Cores from the Normanton Formation and Wallumbilla Formation (part of the Rolling Downs Group (Figure 6-5): the samples are consisting of fine-grained materials with no visible minerals and pore space. • Cores from Gilbert River Formation (Figure 6-6): individual large quartz grains are visible with little cementation (e.g. clay minerals in between quartz grains). Pore spaces in between quartz grains are visible. The mineralogical (XRD) analysis conducted during this assessment on the cores from Dobbyn 1 exploration wells at the southern edge of the Southern Gulf catchments confirmed the lithological variability of the different formations. It showed that the Gilbert River Formation has between 80% and 90% quartz (by weight), with only minor clay minerals and feldspars (Figure 6-11 and Figure 6-12). This confirms that the entire Gilbert River Formation is likely to have significant permeability and is a productive aquifer at this location. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au The Normanton Formation, the shallowest GAB formation in the Southern Gulf catchments, is often described as a partial aquifer (Table 2-1). However, the mineralogical analysis suggests that the Normanton Formation at the Dobbyn 1 location is unlikely to be an aquifer due to the dominance of clay minerals (e.g. smectite, kaolinite and mica) and plagioclase with less than 25% quartz (by weight) (Figure 6-11). Similarly, the Allaru Mudstone, Toolebuc Formation and Wallumbilla Formation, which together with the Normanton Formation comprise the Rolling Downs Group (Figure 2-10), also have low quartz content (generally less than approximately 30% by weight) and a dominance of clay minerals and plagioclase. This confirms that the formations of the Rolling Downs Group are likely to represent aquitards or leaky aquitards at this location. Figure 6-5 Allaru Mudstone cores at Dobbyn 1 exploration well at a depth from approximately 232 to 245 m below ground surface Data source: CSIRO Figure 6-6 Gilbert River Formation core sample at Dobbyn 1 exploration well at a depth of approximately 581 m below ground surface. The inset is a zoom showing coarse quartz grains and open pore space in between quartz grains. Data source: CSIRO Several pieces of broken concrete Description automatically generated with medium confidence Within the Southern Gulf catchments, the GRA does not outcrop and is fully covered by Cenozoic sediments and by several hundred metres of Karumba Basin sediments and the rocks of the Rolling Downs Group (Figure 6-7a). Due to the sparse distribution of logs with reliable stratigraphic and geophysical data, the geometry at the basin margin is not well defined. This means that different representations of aquifer geometry are possible. This includes: • Graphical (cross-section) representations that indicate that the GRA terminates at great depths at its western margin within the Southern Gulf catchments (Figure 2-10). This would indicate that the GRA is confined by large thicknesses of low-permeability sediments throughout the entire Southern Gulf catchments, including close to the margin of the Carpentaria Basin. This would suggest that there is unlikely any recharge to the GRA within the Southern Gulf catchments. • Graphical (cross-section) representations that suggests that the GRA extends further to the basin margin where it may subcrop relatively thin sequences of Cenozoic sediments, which would indicate that there may be recharge to the GRA at the Carpentaria Basin margin within the Southern Gulf catchments. This alternative geometry is shown in Figure 6-7c. Figure 6-7 Three-dimensional geological model of the Southern Gulf catchments showing (a) thicknesses and subsurface geometry of Gilbert River and overlying stratigraphic units; (b) spatial distribution of chloride mass balance recharge estimations; and (c) illustration of alternative subsurface geometry of Gilbert River Formation at the basin margin where the Gilbert River Formation is in direct contact with Cenozoic strata Data sources: Gallant et al. (2011), Department of Environment, Science and Innovation (2021), Vizy and Rollet (2023a) Figure 6-8 Thickness (isopach) map of the Gilbert River Formation Data source: Vizy and Rollet (2023a) Figure 6-9 Depth to the top of the Gilbert River Formation Data sources: Vizy and Rollet (2023a) A map of a large area Description automatically generated Figure 6-10 Cross-section through the Carpentaria Sub-basin of the Great Artesian Basin within the Southern Gulf catchments. See Figure 6-9 for cross-section location Data sources: Gallant et al. (2011), Vizy and Rollet (2023a) Figure 6-11 Mineralogical composition (X-ray diffraction, XRD) at Dobbyn 1 exploration well A diagram of layers of water Description automatically generated For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-12 Mineralogical composition (X-ray-fluorescence, XRF) at Dobbyn 1 exploration well 6.1.3 Alluvial and other aquifers Spatially extensive alluvial aquifer systems occur in association with stream channels, streambeds and floodplains of the middle to lower reaches of the Leichhardt, Nicholson and Gregory rivers, Settlement Creek, and their tributaries within the Southern Gulf catchments (Figure 5-18). Data including lithological, stratigraphic, standing water level and hydrochemical data for these aquifers are generally sparse throughout the Southern Gulf catchments. Block diagrams were developed along two of four existing bore transects across the alluvial plains of the Gregory and Nicholson rivers (Figure 6-13 and Figure 6-14), where the alluvial aquifers are incised into the Cenozoic Karumba Basin. Boundaries between the alluvial aquifers and the underlying Karumba Basin cannot always be clearly differentiated based on lithological logs. At transect 1 near the confluence of the Gregory and Nicholson rivers (Figure 6-13), the thickness of the alluvium ranges from less than 10 m to approximately 25 m. The composition of the alluvial sediments is highly variable: at some of the bores along this transect, coarser sediments such as sand and gravels were intersected, whereas finer sediments such as clay and sandy clay dominate elsewhere. Similar to transect 1, the thickness of alluvial sediments at transect 2 (Figure 6-14), which is located further upgradient within the Gregory River subcatchment, also ranges from less than 10 m to approximately 25 m, and the lithological logs show variable sequences of fine and coarse- grained sediments. Historical water-level measurements collected between 1972 and 1977 are only available for four groundwater bores along these two transects. They suggest that the groundwater table is relatively close to the base of the alluvial aquifers, which indicates that the saturated thickness of 112 | Characterising groundwater resources in the Southern Gulf catchments For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au the alluvium may be relatively thin. However, no screen information is recorded in the groundwater database for these bores. The attribution of the aquifers at the screened interval is therefore highly uncertain and it cannot be determined if the water-level measurements represent the alluvial aquifers or aquifers of the underlying Karumba Basin. Aquifers are also likely hosted in other rocks within the Southern Gulf catchments, including in fractured and weathered Proterozoic igneous rocks and Proterozoic metasedimentary and metamorphic rocks. Recharge rates estimated for some of these rocks north-west of Hells Gate Roadhouse in the northern part of the Southern Gulf catchments are very high (Figure 5-15) and groundwater salinities are very low. However, groundwater data are generally sparse, and aquifers are likely localised and further work would be required to assess the prospectivity of these aquifers. Figure 6-13 Block diagram of alluvium and sub-alluvial bedrock or older Cenozoic sediments along bore transect 1 near the confluence of the Gregory River and Nicholson River. The location of the transect is shown on Figure 6-7 Figure 6-14 Block diagram of the Gregory River alluvium and sub-alluvial older Cenozoic sediments (Karumba Basin) or bedrock along bore transect 2. The location of the transect is shown on Figure 6-7 For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 6.2 Groundwater recharge and flow This section examines groundwater recharge and flow for each stratigraphic formation. The results are first presented for the Cambrian dolostone and limestone in Section 6.2.1, and then for the Gilbert River Formation and Rolling Downs Group in Section 6.2.2. 6.2.1 Cambrian dolostone and limestone As described in Section 6.1.1, the basement topography underneath the Georgina Basin and the underlying South Nicholson Basin is complex, with multiple sub-basins separated by basement highs, including the Alexandria–Wonarah High (Figure 6-2a). Groundwater flow within the Cambrian dolostone and limestone is influenced by the presence of these regional geological structures, and regional groundwater head maps from Knapton et al. (2024) and Dixon-Jain et al. (2024) suggest that at least two major regional groundwater flow divides are present (Figure 6-15). This led to the differentiation of the Georgina Basin aquifer systems into the three distinct Roper River, Lawn Hill Creek and Southern Georgina Basin flow systems (Dixon-Jain, 2024). Within the Southern Gulf catchments, the Georgina Basin is represented by the Lawn Hill Creek flow system (Figure 6-15). In this area, groundwater flow in the Cambrian limestone and dolostone is predominantly towards the east and north-east. Figure 6-15 Three-dimensional representation of regional groundwater heads and inferred groundwater flow lines in the Cambrian dolostone and limestone of the Georgina Basin SWL = standing water level. Data sources: Gallant et al. (2011), Carr et al. (2016), Knapton et al. (2024), Vizy and Rollet (2024) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Groundwater recharge estimates derived using the CMB method have been presented in Section 5.5 In another study assessing recharge to the CLA in the Roper River basin (corresponding to the Roper River flow system), Taylor et al. (2024) suggested that recharge to the Cambrian Limestone Aquifer (CLA) occurs across almost the entire spatial extent of the aquifer where it is unconfined. Based on hydrochemical and environmental tracer data, the authors also suggested that localised recharge occurs via sinkholes (formed by the dissolution of carbonate rocks) near the aquifer outcrop or where surficial features such as sinkholes and streams are incised through the Cenozoic cover. Although no comprehensive mapping of sinkholes exists within the Lawn Hill Creek flow system of the Cambrian dolostone and limestone of the Georgina Basin in the Southern Gulf catchments, a recent study by Dixon-Jain et al. (2024) inferred the presence of sinkholes through the known location of subterranean groundwater-dependent ecosystems (GDEs) and attributed an enhanced recharge potential to this area. Based on stable isotope data and CMB, the same authors also suggested that rapid recharge with estimates of up to approximately 100 mm/year occurs to the Cambrian dolostone and limestone within this area. The CMB recharge assessment conducted in the current study (Section 5.5) estimated slightly lower recharge rates for the Cambrian dolostone and limestone. For the Camooweal Dolostone, estimated recharge ranged from 3.4 to 28 mm/year (with a median of 11 mm/year). For the Thorntonia Limestone, recharge rates ranged from 3.4 to 30 mm/year (with a median of 10 mm/year). However, the spatial distribution of the median CMB recharge estimates (Figure 6-17) shows that there is a considerable spatial variability throughout the Cambrian dolostone and limestone aquifer. The variability is likely related to the spatial variability of presence or absence and variable thickness of Cenozoic cover and the occurrence of karst features, developed through the formation of cavities, sinkholes and fracture networks (Dixon-Jain et al., 2024).The depth of groundwater within the Cambrian dolostone and limestone in the Lawn Hill Creek flow system is variable, and ranges from less than 10 m to more than 100 m (Figure 6-16). Shallow groundwater levels of less than 10 m are observed primarily near the O’Shannassy River, although the data sparsity near the Lawn Hill Creek and Gregory River means that shallow groundwater levels may also occur here. In this area, the confining Wonarah Formation is absent and the carbonate rocks of the Georgina Basin outcrop with little or no Cenozoic cover. Although groundwater-level measurements are sparse in this area, the assessment of potential groundwater discharge areas (Section 5.6.3) indicates that there is a high potential for connectivity between carbonate rock aquifers and streams in the headwaters of the Gregory River and Lawn Hill Creek (Figure 5-18). This is also in agreement with the presence of springs mapped in this area at the eastern edge of the Lawn Hill flow system (Figure 6-15) and spatially extensive important wetlands (Section 2.6; Figure 2-14). Discharge via springs from the Cambrian dolostone and limestone to the headwaters of Gregory River, Lawn Hill Creek, O’Shannassy River at this eastern margin of the Lawn Hill flow system and Georgina Basin may sustain dry-season flows in these areas, as previously also suggested by other authors (Tickell, 2003; Buchanan et al., 2020; Dixon-Jain et al., 2024). Based on interpretation of widely spaced airborne electromagnetic (AEM) data acquired by Geoscience Australia through the AusAEM Stage 1 AEM survey of the NT and western Queensland (Ley-Cooper and Brodie 2018), Dixon-Jain et al. (2024) suggested that discharge to springs and rivers at the eastern margin of the Georgina Basin within the Southern Gulf catchments may also occur via structures that facilitate potential inter-basin connectivity from the Georgina Basin into the Constance Sandstone (a member of the South Nicholson Group) and to springs. However, the authors emphasised knowledge gaps relating to the low spatial resolution of AEM (with a 20 km line spacing, as discussed further in sections 6.3.1 and 7.3 of this report). Furthermore, a common knowledge gap identified by the Southern Gulf Water Resource Assessment and in previous studies (e.g. Buchanan et al., 2020 and Dixon-Jain et al., 2024) is the lack of hydrochemistry and environmental tracer data of springs, streamflow and aquifers within the Lawn Hill Creek flow system of the Georgina Basin. Figure 6-16 Regional depth to water in the Georgina Basin SWL = standing water level. Data source: Knapton et al. (2024) To further assess this, the extent of the outcrop of the Constance Sandstone and the extent of the Georgina Basin in relation to the local topography are also shown in Figure 6-19. This illustration highlights that a significant hydraulic head would be required to facilitate inter-basin discharge from the Georgina Basin through the Constance Sandstone of the South Nicholson Basin to the Boodjamulla Spring Complex (Figure 6-19), as the springs discharge at a surface elevation approximately 50 m higher than the elevation of the ground surface at the eastern margin of the Georgina Basin. This indicates that local recharge – discharge processes (e.g. discharge at the break of slope) within the Constance Sandstone may be the more likely pathway for spring discharge. Evidence of groundwater discharge occurring over multiple timescales in this area is also evident in the observed streamflow data — streamflow being the integrated result of hydrological processes occurring upstream of the point of observation. At two streamflow gauging stations along the Gregory River downstream of the Georgina Basin and Constance Sandstone, the lowest instantaneous streamflow during the month of September (i.e. mid-to-late dry season) between 1970 and 2021 ranged between about 2 m3/s to over 11 m3/s (Figure 6-18). With little rainfall during most dry-season months in the Southern Gulf catchments, lowest instantaneous streamflow during September is likely to be strongly correlated to the volume of groundwater discharging to the Gregory River in most years. Figure 6-18 shows that after a very wet wet-season (indicated by a rapid rise in the cumulative rainfall residual curve) streamflow during the following dry-season is highly elevated and that without considerably higher than average wet season rainfalls during subsequent years, minimum September streamflow declines over a period of about 2 to 5 years. Figure 6-17 Three-dimensional geological model of the Georgina Basin and upscaled median recharge (Section 5.5) within the Southern Gulf catchments Data sources: Gallant et al. (2011), Carr et al. (2016), Knapton et al. (2024), Vizy and Rollet (2024) Figure 6-18 Minimum observed September streamflow at two stream gauge locations on the Gregory River CRR is the cumulative rainfall residual curve. Station 912101A is at the township of Gregory and 912105A is approximately 60 km upstream at Riversleigh (Figure 4-3). No groundwater isotope tracers for dating groundwater were collected as part of this study. Estimating groundwater recharge based on tracer composition or understanding the connectivity between groundwater, surface water and springs was therefore not possible for the Cambrian dolostone and limestone. However, isotope tracers for a small number of groundwater bores within the extent of the carbonate aquifers in the Southern Gulf catchments were presented by Geoscience Australia as part of a Northern Australia Hydrochemical Survey (Schroder et al., 2020), enabling a qualitative assessment of groundwater recharge and flow dynamics. A graph of different colored lines Description automatically generated Figure 6-19 a) Topography (digital elevation model) of western part of the Southern Gulf catchments, streams and mapped springs and b) spatial extent of Constance Sandstone and Georgina Basin with inset cross-section along line A-B Data sources: Raymond et al. (2012), Carr et al. (2016), Gallant et al. (2018), Department of Environment, Science and Innovation (2021) A screenshot of a computer generated image Description automatically generated Stable water isotopes to determine groundwater recharge Due to the inability to go into the field during this study and the paucity of data within the boundary of the Southern Gulf catchments, the Assessment area was extended westwards. This was done in recognition that processes outside the boundary are likely to influence groundwater dynamics within the Southern Gulf catchments. Within the extended Assessment area (including the Undilla Sub-basin), a subset of 35 boreholes screened in the Cambrian limestone and dolostone (sampled by Geoscience Australia from June to August 2019) was used to assess water-stable isotope signatures. The mean, median, minimum and maximum water-stable isotope values from this campaign are displayed in Table 6-1. Table 6-1 Mean, median, minimum and maximum water-stable isotope values for groundwater located in Cambrian dolostone and limestone aquifer in the extended Assessment area For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Oxygen-18 (δ18O) versus deuterium (δ2H) values measured in groundwater within the Cambrian dolostone and limestone in the extended Assessment area are shown in Figure 6-20 . These data are plotted with the amount-weighted mean water-stable isotopes in rainfall measured at Mount Isa (δ18O = –6.96‰ and δ2H = –44.7‰) to define the local meteoric water line (LMWL) as δ2H = – 7.73 × δ18O + 9.10 (Hollins et al., 2018). The global meteoric water line (GMWL) δ2H = 8 × δ18O + 10 from Craig (1961) is also included in the plot. Figure 6-20 Water-stable isotopes measured (δ2H vs δ18O) in groundwater located in Cambrian dolostone and limestone aquifer (adopted from Dixon-Jain et al., 2024) VSMOW = Vienna standard mean ocean water; LMWL = local meteoric water line; GMWL = global meteoric water line. Data sources: Hollins et al. (2018), Schroder et al. (2020) Most samples plot around the 100 to 200 mm/month rainfall mean, indicating that recharge is predominantly from rainfall events greater than 100 mm/month, which is approximately a 1 in a 10-year event (Dixon-Jain et al., 2024). A few samples in particular plot to the right of the Mount Isa LMWL, showing a clear evaporation signature, which indicates evaporation occurred prior to recharge at these locations. The stable water isotopes measured in the Cambrian dolostone and limestones are also displayed spatially in Figure 6-21. In that figure, blue dots and squares show samples with relatively high δ18O and δ2H values, which are more enriched in oxygen-18 and deuterium, compared to the red dots or squares with lower δ18O and δ2H values. One borehole in the western part of the study area (89 m depth; RN018425) displayed the most evaporated values (enriched in δ18O and δ2H), showing that evaporation occurred prior to recharge. Samples near this site, to the south-west of the groundwater flow divide, generally have lower δ18O and δ2H values (between –8.34‰ and – 3.39‰ for δ18O and between –59.5‰ and –33.99‰ for δ2H) compared to the samples to the north-east, supporting the presence of the groundwater divide in this area. Moreover, the slightly higher δ18O and δ2H values to the north-east of the flow divide are possibly due to recharge from smaller monthly average rainfall events (e.g. 50 to 100 mm/month) due to the aquifer outcropping in this area. A graph of a graph showing the amount of rainfall Description automatically generated with medium confidence Figure 6-21 Water-stable isotopes measured in groundwater. Low values are displayed in red while high values are displayed in blue Data source: Schroder et al. (2020) Radiocarbon composition to determine groundwater recharge The interpretation of radiocarbon (14C) values measured in groundwater in the Cambrian limestone and dolostone poses significant challenges due to the addition of dead carbon from either dissolution of carbonate during recharge and/or isotopic exchange during flow through the carbonate aquifers. Addition of dead carbon leads to calculated groundwater radiocarbon ages that are significantly older than the actual age. Incorporation of dead carbon necessitates careful correction and calibration of the data to obtain accurate radiocarbon dates. Despite these complexities the uncorrected radiocarbon data can provide qualitative but valuable preliminary insights into hydrogeological processes occurring in the Cambrian dolostone and limestone aquifer. Radiocarbon is measured as the percentage of modern carbon (pMC). That means the higher values represent relatively young groundwater and lower values represent old groundwaters of up to approximately 40,000 years. The comparison of groundwater signatures in carbon-14 (pMC) and carbon-13 (per mil) provides insights into the origin and age of groundwater. Carbon-14, through its radioactive decay, is used for dating, while stable carbon-13 helps identify geochemical processes like carbonate dissolution (e.g. –2‰ to zero ‰) or interaction with organic matter (e.g. –20‰ to –10‰). Values within the Undilla Sub-basin range from 5.58 to 96.21 pMC with a carbon- 13 varying from –14.23‰ to –6.21‰ (Figure 6-22). Figure 6-22 Carbon-13 vs carbon-14 in groundwater under the Undilla Sub-basin VPDB = Vienna PeeDee Belemnite; pMC = percent modern carbon. Data source: Schroder et al. (2020) Eleven radiocarbon samples in Figure 6-22 and Figure 6-23 show that the dataset comprises two samples lower than 10 pMC, four samples between 10 and 25 pMC, three samples between 25 and 75 pMC and two samples with radiocarbon higher than 75 pMC (Figure 6-23). Out of ten measurements within the Assessment area, only two samples had values greater than 75 pMC, indicating young groundwaters and rapid recharge. These are in the south-western part of the Southern Gulf catchments (Figure 6-23) where aquifer outcrops allow rapid recharge. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-23 Radiocarbon value in percent modern carbon (pMC) of the Cambrian dolostone and limestone aquifer displayed with the simplified geology as background Data source: Schroder et al. (2020) 6.2.2 Gilbert River Formation and Rolling Downs Group Potentiometric surface and groundwater flow directions for the Gilbert River Formation and wider GAB area were updated in 2015 by Ransley et al. (2015) (Figure 6-24). There has not been sufficient water-level data collected to update these, so they have been used as published by Ransley et al. (2015) in this study. Generally, groundwater flow in the Gilbert River Formation is from the north-east boundary towards the northern coast or from the south-east outside the study area towards the northern coast. For selected bores across the Southern Gulf catchments, hydrographs are shown for the Gilbert River Formation in Figure 6-25. While there can be large drawdowns, especially in the eastern parts of the GAB due to the drilling and discharge of artesian groundwater (Ransley and Smerdon, 2012), water levels are relatively stable within the Southern Gulf catchments. Figure 6-24 Great Artesian Basin (GAB) potentiometric surface and generalised flow lines GAB potentiometric surface, contour lines, inferred flow directions from Ransley et al. (2015). The potentiometric surface raster does not extend to the full extent of the Gilbert River Formation as some of the GAB bores with hydrochemistry data from this study (plotted in grey) are outside the raster extent. Location of Gilbert River Formation bores in Figure 6-25 shown in yellow. Data source: Ransley et al. (2015) Figure 6-25 Selected groundwater-level hydrographs from the Gilbert River Formation Location of Gilbert River Formation bores shown in Figure 6-24. Water-level data from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023). 6.3 Groundwater salinity and hydrochemistry by stratigraphic formation This section examines the salinity and hydrochemistry results for each stratigraphic formation. The results are first presented for the Cambrian dolostone and limestone in Section 6.3.1, and then for the Gilbert River Formation and Rolling Downs Group in Section 6.3.2. 6.3.1 Cambrian dolostone and limestone Groundwater chemistry and its implication for irrigation water quality To assess groundwater chemistry, the followings graphs illustrate the relationship between chloride (Cl⁻) concentrations and various major ions (Ca²⁺, Mg²⁺, Na⁺, HCO₃⁻, SO₄²⁻, Fe and SiO₂) in the Cambrian dolostone and limestone aquifer within the Undilla Sub-basin. The electrical conductivity (EC) is displayed as a colour gradient while the dashed lines represent the seawater dilution line, indicating the expected ion concentrations if the groundwater were diluted with seawater (Figure 6-26). The data show that while some samples closely follow the seawater dilution line, particularly for Na⁺, others deviate, suggesting that different geochemical processes, such as mineral dissolution, ion exchange or anthropogenic influences, underlie their concentration. The positive correlation between element concentration and EC indicates that mineralisation and salinity are acquired along the flow path within the aquifer. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-26 Relationship between chloride (Cl⁻) concentrations and various major ions (Ca²⁺, Mg²⁺, Na⁺, HCO₃⁻, SO₄²⁻, Fe and SiO₂) EC = electrical conductivity. Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Determining the suitability of irrigation water salinity for a crop Geochemical data from 226 boreholes completed in the Cambrian dolostone and limestone were selected within the Undilla Sub-basin. According to the Australian Water Resources Council’s saline water quality classes (AWRC, 1976), most of the groundwater resources in the Undilla Sub- basin fall within the fresh to marginal water quality classes, with mean TDS of 1096 mg/L (median = 1040 mg/L) (Figure 6-27). Very fresh water appears to be located in the northern Undilla Sub- basin (coinciding with the upper reaches of the Gregory River and Lawn Hill Creek), with a trend from low salinity (from 244 mg/L) within and near the groundwater flow divides to more saline water (reaching 6112 mg/L) towards the south and south-west of the boundary of the Undilla Sub- basin (Figure 6-27). This increase in salinity tends to follow the expected groundwater flow direction as the TDS increases due to water–rock interactions along the groundwater flow paths (Figure 6-27). The density of available chemistry data is not uniformly distributed in the Cambrian dolostone and limestone, with lower density of data in Queensland than in the NT (Figure 6-27). Figure 6-27 Map displaying salinity by total dissolved solids (TDS) in milligram per litre values within the model survey area plotted with simplified geological units The sodium (Na) and chloride (Cl) concentrations are plotted in Figure 6-28 showing the concentrations that can cause foliar injury in crops. The sodium concentration reaches 1270 mg/L, with mean and median concentration of 155 and 124 mg/L respectively. The chloride concentration reaches 2450 mg/L, with a mean and median concentration of 254 and 130 mg/L, respectively. Most of the groundwater is suitable for moderately tolerant crops (N = 242). According to the tolerance of crops to salinity in irrigation water detailed in the ANZECC and ARMCANZ (2000) classification, the mean groundwater salinities are good for supply to most field crops and fruits (except avocado (Persea americana) and apple (Malus spp.)), pasture (except clover (Trifolium spp.)) and vegetables (except pepper (Capsicum spp.), lettuce (Lactuca spp.), onion (Allium cepa), beans (family: Fabaceae) and carrot (Daucus carota)) for clay soils. The mean groundwater salinities are good for supply to all field crops, fruits, pasture and vegetables in loam and sandy soils (Figure 6-28). However, some groundwater samples (N = 13) with a chloride concentration greater than 700 mg/L can cause foliar injuries to these same crops. As detailed in the ANZECC and ARMCANZ (2000), the crops concerned are: • almond, apricot, citrus, plum, grape as sensible; • pepper, potato (Solanum tuberosum), tomato as moderately sensitive; • barley, maize, cucumber, lucerne as moderately tolerant, and; • cauliflower, cotton, sugar beet, sunflower as tolerant). The highest chloride concentrations are observed along the western boundary of the Undilla Sub-basin survey area (Figure 6-29). Irrigation water must therefore be chosen carefully so as not to risk damaging the crops, especially along the north-east to south-west boundary of the model survey area. Figure 6-28 Sodium (Na) against chloride (Cl) in the Cambrian dolostone and limestone in the Undilla Sub-basin area Halite dissolution line is displayed in grey, seawater dilution line in red, and dotted black lines represent different threshold lines of plant tolerance. TDS = total dissolved solids. Figure 6-29 Chloride (Cl) concentration in groundwater (milligrams per litre) in Undilla Sub-basin displayed by crop suitability to concentration sensitivity, plotted with simplified geological units Crops concerned – Sensitive: almond, apricot, citrus, plum, grape; Moderately sensitive: pepper, potato, tomato; Moderately tolerant: barley, maize, cucumber, lucerne; Tolerant: cauliflower, cotton, sugar beet, sunflower. Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) According to ANZECC and ARMCANZ (2000), no adverse effects on livestock are expected if the concentration of sulfate in drinking water does not exceed 1000 mg/L. Adverse effects may occur at concentrations between 1000 and 2000 mg/L (especially in young or lactating animals or in dry, hot weather) and may be transient and may disappear as animals become accustomed to the water. Sulfate levels above 2000 mg/L may cause chronic or acute health problems in livestock. The majority of groundwaters studied here present a sulfate concentration largely below the threshold of 1000 mg/L, and so will present no adverse effects to stock. The highest concentration in sulfate is found in bores in the north-west of the survey area, as displayed in Figure 6-30. Figure 6-30 Map of sulfate (SO4) concentration measure in Cambrian limestone and dolostone groundwater in Undilla Sub-basin (milligrams per litre) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) The mean pH is 7.5 ±0.5 so there is limited corrosion potential to pumping, irrigation and stock watering equipment from these waters (ANZECC and ARMCANZ, 2000). There is insufficient data to investigate if there are any nutrients, pesticides, metals or radioactive contaminants in the Cambrian dolostone and limestone in this area. Determining the risk of soil structure degradation caused by irrigation water quality Another way of estimating the quality of groundwater use for irrigation is to compare the sodium adsorption ratio (SAR) with EC to predict the stability of soil structure in relation to irrigation water. The SAR value measures the relative concentration of sodium (Na+) to calcium (Ca2+) and magnesium (Mg2+) in millimole per litre as calculated in ANZECC and ARMCANZ (2000). The mapped SAR data align with the chloride data (Figure 6-29) showing a trend from low in the north-east to high towards the south-west of the survey area (Figure 6-31). The SAR values range up to 29 mmol/L, with mean and median concentrations of 4 and 3 mmol/L respectively. Figure 6-32 shows the SAR values plotted again EC, while the colour of the data points indicates values in TDS. The figure also displays threshold lines for two soils with different clay content (line A: 55% to 65% and line B: 25% to 35%) and mineralogy for an annual rainfall of 1000 mm/year. The soils are unstable in the areas to the left of the blue line and increasingly stable to the right of the red line. The most inadequate SAR values are found in isolated areas, one in the north-east of the Undilla Sub-basin and the other in the south-west (Figure 6-31). For the remainder of the boreholes shown in this study, the SAR value indicates that irrigation with these groundwaters will be conservative of the soil structure (Figure 6-32). Figure 6-31 Map of sodium adsorption ratio (SAR) concentration in millimole per litre displayed with the simplified geology in background Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-32 Sodium adsorption ratio (SAR) and electrical conductivity (EC) calculated where colour represents total dissolved solids (TDS), the stability of soil is highlighted by two threshold lines displayed for soils composed by different clay concentrations (Line A: 55% to 65% and Line B: 25% to 35%) for an annual rainfall of 1000 mm/year Data source: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Groundwater chemistry and major mineral weathering processes in groundwater The relative proportions of major anions (chloride (Cl) and bicarbonate (HCO3)), and cations (calcium (Ca) and magnesium (Mg) to sodium (Na) and potassium (K)) of each groundwater sample from the Cambrian dolostones and limestones are plotted on a Piper diagram to show their hydrochemical facies (Figure 6-33). These groundwaters appear to be the result of mixing or more likely geochemical evolution along the groundwater flow path as they show a wide range of water types, which is correlated with an increase in TDS. Waters with low to medium TDS show a magnesium bicarbonate type (Mg–HCO3), while water with medium to high TDS is sodium chloride type (Na–Cl). Several samples also show an Mg–Cl and Ca–HCO3 type, the latter normally expected for limestone weathering processes, while the Mg–HCO3 type is expected for dolostone weathering processes. Indeed, limestone dissolution increases the calcium and bicarbonate content, while dolomite dissolution would also increase the magnesium content in groundwater. The major mineral weathering processes controlling the composition of the groundwaters are discussed below. Figure 6-33 Piper diagram of Cambrian limestone and dolostone groundwater displaying only samples with an ionic balance less than 10% and a complete dataset (N = 213). The colour represents the associated total dissolved solids (TDS) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Solute concentrations in the Cambrian dolostones and limestones throughout the study area vary considerably: Cl 2 to 2450 mg/L, alkalinity (as HCO3) 5 to 731 mg/L, SO4 2 to 1620 mg/L, Na 7 to 1270 mg/L, Ca 14 to 412 mg/L, Mg 11.4 to 362 mg/L and K 1 to 105 mg/L. The Chandha diagram (Figure 6-34) highlights the weathering processes along the flow paths within the Cambrian units. It shows that the water facies are modified along water flow paths by an increase of EC from fresh water of Ca–Mg–HCO3 water type to saline water of Na–Cl water type. Figure 6-34 Chadha diagram in percent milliequivalents per litre plotted where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) To investigate the processes controlling the major solute concentrations in groundwater along the flow path, the concentration in milligrams per litre or milliequivalents per litre has been plotted in several plots below (from Figure 6-35 to Figure 6-37). The use of milliequivalents per litre units in these plots allows for identification of the dissolution processes (through dissolution lines), which are necessary to identifying weathering processes. Figure 6-35 shows calcium (Ca) and magnesium (Mg) plotted against bicarbonate in milliequivalents per litre, allowing us to specify the origin of calcium and magnesium. It shows that Ca and Mg concentrations observed in fresh to marginal groundwaters generally plot along or above the dolomite dissolution line, thus indicating dolomite dissolution to be a major mineral weathering process. While not obvious from the plot, it is likely calcite dissolution also occurs, considering the prevalence of calcite rocks in addition to dolomite. Marginal to saline groundwaters, indicated by a higher EC, appear to have another Ca or Mg input. Figure 6-35 Ca and Mg versus bicarbonate concentration in milliequivalents per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Figure 6-36 shows calcium (Ca) plotted against sulfate (SO4) concentration in milliequivalents per litre. At lower EC (4000 μS/cm), correlation between concentrations in Ca and SO4 indicates a possible dissolution of gypsum or anhydrite if present. However, there is no clear correlation between the Ca and SO4 concentration in saline groundwaters (EC >4000 μS/cm). Thus, the Ca (and SO4) increase in the most saline samples are not due to gypsum dissolution. As the samples generally plot above the carbonate (Figure 6-35) and gypsum dissolution lines (Figure 6-36), this indicates an excess of Ca ions that can only be explained by both carbonates rocks(dolomite and calcite) being dissolved into the groundwater. Marginal to saline groundwaters, with the highest EC, seem to have a different SO4 input. Alternatively, there may be some Ca (and Mg) being deposited and hence lost from solution at these higher salinity samples, as these samples are generally oversaturated with respect to calcium and magnesium (see Figure 6-36). Figure 6-36 Ca versus SO4 concentration in milliequivalents per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) To discriminate groundwaters affected by an increase in SO4 concentration, Figure 6-37 displays (Ca + Mg) – (Na + K) in percent milliequivalents per litre, according to SO4concentration in milligrams per litre and also displays EC values. This figure shows a trend highlighting that fresh water is mainly composed of Ca and/or Mg cations with a low SO4 content, and as the electrical conductivity increases, the Ca–Mg concentration decreases as the relative Na–K content increases. The Na–K increase is likely a result of sodium (Na) increase from halite dissolution, as halite is known to occur in the Anthony Lagoon Formation (Tickell, 2003; Dixon-Jain et al., 2024). The highest SO4 concentrations are found in saline groundwaters. Figure 6-37 (Ca + Mg) – (Na + K) in percent milliequivalents per litre versus sulfate concentration in milligrams per litre where colour represents electrical conductivity (EC) for Cambrian dolostone and limestone data in the Undilla Sub- basin (N = 225) Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) To help specify geochemical processes occurring along flow paths, the saturation indices of certain minerals are calculated and plotted in Figure 6-38. These saturation indices, determined from dissolved element concentrations, indicate the saturation state of each mineral. If the index is positive, the solution is oversaturated, meaning precipitation of the mineral can occur. Conversely, if the index is negative, the solution is undersaturated, suggesting that the mineral is likely to dissolve. Figure 6-38 shows the multiple saturation indices of the following minerals: gypsum (A), anhydrite (B), dolomite (C), aragonite (D) and calcite (E) according to chloride in milligrams per litre. Of the 225 samples studied, no water was saturated with anhydrite but 2 were saturated with gypsum, 178 with dolomite, 137 with aragonite and 161 with calcite. The higher the salinity, the more the groundwater tends to be saturated in gypsum and anhydrite. However, there is no apparent relationship between salinity and saturation in calcite, dolomite or aragonite. Only some fresh groundwaters are undersaturated in calcite, dolomite or aragonite. The maps of the saturation indices (Figure 6-39 and Figure 6-40) highlight that there is a similar distribution in groundwater saturation indices for dolomite (Figure 6-39 and calcite (Figure 6-40). For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-38 Saturation indices of minerals (A: gypsum, B: anhydrite, C: dolomite, D: aragonite, E: calcite) according to chloride (Cl) concentration in milligrams per litre, where colour represents the electrical conductivity for Cambrian dolostone and limestone data in the Undilla Sub-basin (N = 225) Figure 6-39 Mapping of dolomite saturation indices (SI) Figure 6-40 Mapping of calcite saturation indices (SI) The major mineral weathering processes that control the relative increase of calcium (Ca) and magnesium (Mg), as well as alkalinity, is carbonate dissolution (dolomite and calcite). Dissolution of carbon dioxide (CO2) in the soil zone initially drives the dissolution of carbonate minerals, increasing bicarbonate (HCO3), Mg and Ca concentrations. This causes the groundwaters to be predominantly oversaturated with respect to calcite and especially dolomite from a low salinity and lead to the groundwaters being predominantly Mg–HCO3 type at low salinity. Along the flow paths, as TDS increases the addition of Na, HCO3 and chloride (Cl) along with Ca and Mg ions are consistent with dissolution of gypsum, and possibly anhydrite if present, and halite, in addition to further carbonate dissolution. This causes the groundwaters to become oversaturated with respect to gypsum as salinity increases and leads to the groundwaters tending towards being Na– Cl type at high salinities. There may also be some mixing with formations above and below, which are not identified here. Nevertheless, the mineral weathering processes outlined above have also been identified as those controlling the major solute concentrations in the wider South Nicholson and Georgina basins (Dixon-Jain et al., 2024). 6.3.2 Gilbert River Formation and Rolling Downs Group Groundwater chemistry of concern for irrigation water quality In the Gilbert River Formation and Rolling Downs Group, the salinity ranges from 267 up to 28,500 μS/cm (Figure 6-41). On average, salinity is 1915 μS/cm (median 1854 μS/cm) in the Gilbert River Formation and 2365 μS/cm (median 1350 μS/cm) in the Rolling Downs Group within the Southern Gulf catchments (Figure 6-41). These mean salinities are too high for several field crops, pastures, fruits and vegetables grown in loam or clay soils, such as rice (Oryza sp.), avocado or potatoes, but would be sufficient for almost all in sandy soils (ANZECC and ARMCANZ, 2000). Figure 6-41 Map of salinity in Gilbert River Formation and Rolling Downs Group Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) The sodium (Na) and chloride (Cl) concentrations are plotted in Figure 6-42 showing the concentrations that can cause foliar injury in crops. The concentrations are sufficiently high to cause foliar injury in chloride and sodium sensitive, moderately sensitive and tolerant crops, meaning only crops tolerant to chloride and sodium (i.e. cauliflower and cotton) (ANZECC and ARMCANZ, 2000) could be grown using GAB groundwater from a significant portion of the Southern Gulf Water Resource Assessment study area. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-42 Sodium (Na) against chloride (Cl) in the Gilbert River Formation and Rolling Downs Group with seawater dilution line in orange Figure excluding Rolling Downs Group bores RN036255, RN012811 and RN072321 with Cl concentrations of 2150, 2493 and 10,500 mg/L, respectively. Showing sodium and chloride concentrations causing foliar injury in sensitive crops up to dotted black line (Cl and Na <5 mmol/L), moderately sensitive up to dot–dash black line (Cl and Na between 5 and 10 mmol/L), moderately tolerant up to the dashed black line (Cl and Na between 10 and 20 mmol/L) and tolerant beyond the dashed black line (Cl and Na >20 mmol/L; ANZECC and ARMCANZ, 2000). Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Within the Southern Gulf catchments, the Gilbert River Formation dips steeply in the subsurface to depths greater than 700 mBGL towards the Gulf of Carpentaria (Figure 6-43). The chloride concentrations in several bores throughout the centre of the area would be low enough to grow moderately sensitive crops (e.g. potato or tomato) (ANZECC and ARMCANZ, 2000). However, in the remaining bores, the chloride concentration is sufficiently high that only tolerant or moderately tolerant crops (e.g. barley and cauliflower, respectively) could be grown. Only those bores along the margin of the GAB (extent of the Cretaceous sediments; Figure 6-43) have relatively shallow depths between 28 and approximately 350 mBGL and chloride concentration low enough to grow sensitive crops (e.g. citrus) (ANZECC and ARMCANZ, 2000). Likewise, those bores outside the Southern Gulf catchments, west of Cloncurry (Figure 6-43), have depths less than 350 mBGL and chloride concentration low enough to grow sensitive crops. A map of the united states Description automatically generated Figure 6-43 Map of depth to the top of the Gilbert River Formation and total dissolved solids (TDS) in the Gilbert River Formation Data sources: TDS from the NTG Data Portal (Department of Environment Parks and Water Security, 2014) and from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Almost all of the Rolling Downs Group bores are less than 250 m deep (Figure 6-44). The chloride concentration in most of these bores would be low enough to grow sensitive crops (e.g. citrus) (ANZECC and ARMCANZ, 2000). However, in several bores towards the Gulf of Carpentaria the chloride concentration is higher meaning that sensitive crops and moderately tolerant crops (e.g. tomato and barley, respectively) could not be grown using the water. Also, these bores are predominantly completed in the Wallumbilla Formation (Table 6-4), which is considered a partial aquifer, so long-term yields may be low. A map of the united states Description automatically generated Figure 6-44 Map of the depth to the top of the Gilbert River Formation, bore depth to the top of the Rolling Downs Group and total dissolved solids (TDS) in the Rolling Downs Group Data sources: TDS and Rolling Downs Group depth from the NTG Data Portal (Department of Environment Parks and Water Security, 2014) and from the Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) The groundwater mean pH is 8.2 (range from 6.9 to 9) so there is limited corrosion potential to pumping, irrigation and stock watering equipment from these waters (ANZECC and ARMCANZ, 2000). There is insufficient data to investigate if there are any nutrients, pesticides, metals or radioactive contaminants in the GAB in this area. Groundwater chemistry and major mineral weathering processes The water samples are plotted on a Piper diagram to show the hydrochemical facies (Figure 6-45). Groundwaters are mostly of the Na–HCO3 type (Figure 6-45) in the Gilbert River Formation (82%) and Rolling Downs Group (62%) (Table 6-2). A number of samples are also of the Na–Cl type in the Gilbert River Formation (18%) and Rolling Downs Group (34%), and a single Rolling Downs Group sample is of the Ca–SO4 type (Figure 6-45). The major mineral weathering processes controlling the composition of the groundwaters are discussed below. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-45 Piper diagram of groundwater samples Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Solute concentrations in the GAB groundwaters vary considerably throughout the study area. Measured solute concentrations in the Gilbert River Formation also vary: Cl 16 to 3300 mg/L, alkalinity (as HCO3) 148 to 960 mg/L, SO4 1 to 700 mg/L, Na 81 to 990 mg/L, Ca 0.4 to 553 mg/L, Mg 0.1 to 630 mg/L and K 0.8 to 62 mg/L (Figure 6-42 and Figure 6-46). The solute concentrations in the Rolling Downs Group are similar, varying between: Cl 20 to 10,500 mg/L, alkalinity (as HCO3) 89 to 895 mg/L, SO4 1 to 1550 mg/L, Na 38 to 6300 mg/L, Ca 1 to 440 mg/L, Mg 0.1 to 460 mg/L and K 1 to 61 mg/L (Figure 6-42 and Figure 6-46). To investigate the processes controlling the major solute concentrations, they are plotted against chloride in Figure 6-42 and Figure 6-46. These show that calcium and sulfate concentrations correlate with chloride concentrations to a large degree, with samples generally plotting along the seawater dilution line (Figure 6-46). Thus, it is likely that evapoconcentration is a major process controlling these solute concentrations along the groundwater flow paths. Nonetheless, mixing with more saline water, and/or diffusion from the aquitard, could also be contributing to a small degree. Table 6-2 Number of bores of the Na–HCO3, Na–Cl and Na–SO4 types in the Gilbert River Formation and Rolling Downs Group For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au A screenshot of a graph Description automatically generated Figure 6-46 (a) Ca, (b) Mg, (c) HCO3 and (d) SO4 against chloride (Cl) concentration in the Gilbert River Formation and Rolling Downs Group Figure excluding Rolling Downs Group bores RN036255, RN012811 and RN072321 with Cl concentrations of 2,150, 2,493 and 10,500 mg/L, respectively. Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) The importance of mineral weathering is also highlighted for a number of the major solutes. This is shown with sodium concentrations and alkalinity (as HCO3) plotting above the dilute seawater line (Figure 6-42 and Figure 6-46), and below the seawater dilution line for magnesium concentrations (Figure 6-46). The major mineral weathering processes that control the relative increase of sodium and alkalinity and comparative reduction in magnesium (and calcium) is carbonate dissolution and cation exchange. Dissolution of CO2 in the soil zone initially drives the dissolution of carbonate minerals, increasing HCO3 and Ca concentrations. This causes the groundwaters to be predominantly oversaturated with respect to calcite and especially dolomite (Figure 6-47). This is then followed by cation exchange of sodium for calcium and magnesium, reducing magnesium and calcium concentrations and increasing sodium concentrations in groundwater. It is these processes that lead to the groundwaters being predominantly Na–HCO3 type (Figure 6-45). These mineral weathering processes have also been identified as those controlling the major solute concentrations in the wider GAB (Herczeg et al., 1991; Moya et al., 2015; Priestley et al., 2020; Radke et al., 2000). A screenshot of a graph Description automatically generated Figure 6-47 (a) Calcite and (b) dolomite saturation indices against calcium (Ca) concentration in the Gilbert River Formation and Rolling Downs Group Figure excluding Rolling Downs Group bores RN036255, RN012811 and RN072321 with Cl concentrations of 2,150, 2,493 and 10,500 mg/L, respectively. The GAB groundwaters change from Na–HCO3 type to Na–Cl type waters as the chloride concentration increases (>200 mg/L) (Figure 6-48). This change in water type is due to increasing salinity and solute concentrations, such as Cl (Figure 6-48a), rather than a loss of HCO3 (Figure 6-48b). Therefore, this change from bicarbonate dominant waters into Na–Cl type waters with increasing salinity along the groundwater flow path is likely due to mixing with more saline water along the groundwater flow path or diffusion of solutes, such as Cl, Na, Ca, Mg and SO4, from the aquitards of the Rolling Downs Group (e.g. Priestley et al., 2020; Radke et al., 2000). The single Ca–SO4 type water was from the Rolling Downs Group (Figure 6-45), located east of Cloncurry. It is likely caused by localised dissolution of gypsum in the sedimentary units because the surrounding Rolling Downs Group are of Na–HCO3 type (Doan, 1988; Krieg et al., 1995). A graph of different colored dots Description automatically generated with medium confidence Figure 6-48 (a) Na and (b) HCO3 against chloride (Cl) concentration in the Gilbert River Formation and Rolling Downs Group plotted showing the water type Figure excluding Rolling Downs Group bores RN036255, RN012811 and RN072321 with Cl concentrations of 2,150, 2,493 and 10,500 mg/L, respectively. Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) Other environmental tracer and isotope data are lacking in this area and are still a key data gap within the GAB, as also identified for the larger Carpentaria Basin (Raiber et al., 2022). Of the available data, there were only two carbon isotope results available for GAB bores south-west of Cloncurry (Table 6-3). The δ13C values less than –20‰ indicate there has been some carbonate dissolution. This is also supported by the groundwaters predominantly being of the Na–HCO3 type and oversaturated with respect to calcite and dolomite (Figure 6-47). The 14C results show the groundwaters are relatively old and could have been recharged approximately 30,000 years ago, although presence of dead carbon in the system could reduce the apparent groundwater age significantly. Table 6-3 Geosciences Australia sample number, bore registration number, bore name, longitude, latitude and carbon isotope results for two Gilbert River Formation bores For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au †pMC = percent modern carbon Hydrochemistry along a flow path in the Gilbert River Formation There are sufficient samples along the groundwater flow path south of Burketown (Figure 6-49) to look at the hydrochemistry along a flow line for the Gilbert River Formation. The major solutes are plotted along this flow line in Figure 6-50 and Figure 6-51. Except for one outlier, the chloride, sodium and bicarbonate concentrations increase along the flow path. And conversely, the calcium and magnesium concentrations decrease along the groundwater flow line, while the sulfate concentrations are relatively stable. This supports the mineral weathering processes identified in the regional hydrochemistry figures, namely that evapoconcentration, carbonate dissolution and cation exchange are controlling the relative increase of chloride, sodium and alkalinity and comparative reduction in magnesium (and calcium). A map of a weather forecast Description automatically generated Figure 6-49 Great Artesian Basin (GAB) potentiometric surface and generalised flow line GAB potentiometric surface, contour lines, inferred flow directions from Ransley et al. (2015). The generalised flow line with 30 km buffer zone shows which bores were used to represent groundwater chemistry along a flow line (Figure 6-50 and Figure 6-51). Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-50 Chloride (Cl), HCO3 and SO4 against distance along the generalised flow line using a 30 km buffer zone in Figure 6-49 Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Figure 6-51 Sodium (Na), calcium (Ca) and magnesium (Mg) against distance along the generalised flow line using a 30 km buffer zone in Figure 6-49 Data sources: NTG Data Portal (Department of Environment Parks and Water Security, 2014) and Queensland groundwater database (Department of Regional Development, Manufacturing and Water, 2023) The formations of the Rolling Downs Group As mentioned above, the groundwater chemistry of the Rolling Downs Group is similar in range and spatial variability to that of the Gilbert River Formation (Figure 6-42 and Figure 6-46). The Rolling Downs Group bores are predominantly completed in the Wallumbilla Formation (or its equivalent; 20 out of 29) (Table 6-4), which conformably overlies the Gilbert River Formation (Orr et al., 2020; Smerdon et al., 2012). Hence it is possible that these are hydraulically connected and would explain the similarity in hydrochemistry between the two formations. No bores were identified as completed in the Normanton Formation (Table 6-4) despite it often being referred to as a partial aquifer. Table 6-4 Stratigraphic unit of the Rolling Downs Group bores For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 6.4 Numerical flow modelling The groundwater model of the Undilla Sub-basin developed by Knapton et al. (2024) consists of a two-dimensional finite element model developed using FEFLOW (Section 4.4). The interaction between groundwater and surface water occurs using the FEFLOW-specified head boundary conditions (i.e. 1st type Dirichlet). The FEFLOW groundwater model represents the CLA in the Undilla Sub-basin, which extends to the west beyond the Lawn Hill Creek flow system of the Georgina Basin in the Southern Gulf catchments (Figure 6-15). The Undilla Sub-basin model domain encompasses an area of approximately 50,900 km2 and is referred to as Undilla1 (Undilla Sub-basin groundwater flow model v1). The Undilla1 groundwater model was developed with all available aquifer data, and was calibrated with all available rainfall, river flow and groundwater-level data. The recharge inputs to the FEFLOW model were generated by scaling the CMB estimates of recharge (Crosbie and Rachakonda, 2021). A two-dimensional numerical groundwater flow model has been developed to examine the groundwater resources of the Undilla Sub-basin, which provides baseflow to Lawn Hill Creek and Gregory River. The model successfully reproduced the observed behaviour of groundwater levels and discharge from the CLA aquifer in the Undilla Sub-basin. From this study, the following key findings have emerged: •The conceptualisation of the groundwater flow system indicates that there is a localised systemdischarging to springs well above the stream level and a regional groundwater systemdischarging to lower springs and through the bed of the river. To adequately model theobserved discharge record, the Undilla Sub-basin may require multiple layers to resolve thispartitioning. •There is considerable uncertainty in the dynamic range of groundwater levels in the Undilla Sub- basin, as the model is currently constrained by single water levels recorded at the time of bore construction. Collecting time series data at sites such as the reporting sites used in this study would reduce the uncertainty in the groundwater-level dynamics. •Groundwater discharge reported at stream gauging site 912105A (Figure 4-3) is consideredrepresentative of flows in the Gregory River; however, there is less confidence in the dischargereported at 912103A being representative of flows in Lawn Hill Creek. Conducting manualmeasurements of stream flows at these sites once or twice a year during the dry season wouldimprove confidence in this data. •Portions of the observed flow record can be reproduced for bore 91213A on Lawn Hill Creek. However, the period from 1975 to 1980 appears to underreport flows at 912103A. •Previous studies have assumed that groundwater contributing to the discharge at Lawn HillCreek and Gregory River is sourced as far west as the Alexandria–Wonarah High (Figure 6-2). However, the groundwater head surface indicates that there is a groundwater divide separatingflows to the east and flows to the south (Figure 6-16). This assessment is supported by thegroundwater flow model. •Recharge is estimated by scaling the rainfall by the CMB recharge distribution and is between 7and 25 mm/year, depending on the area of interest. Recharge estimates in the catchment of theGregory River is 14 mm/year, and the Lawn Hill catchment recharge is 25 mm/year. NicholsonGroundwater Management Area (NGMA) estimated recharge is about 15 mm/year, whereas themean recharge for the entire model domain is about 7 mm/year. •Based on the transmissivity values, the recharge for areas with black soil cover may be an orderof magnitude lower. •The transmissivities in the north-eastern third of the model domain are considered reasonablefor the type of aquifer (<1,000 to 10,000 m²/day). However, the highest transmissivity values(>20,000 m²/day), which are predominantly in the south-western two-thirds of the modeldomain, are much higher than expected. The higher values reflect the very low groundwatergradient and are likely a result of recharge being overestimated in areas with black soil. Current assumptions and limitations of the Undilla1 groundwater flow model are: •Groundwater levels are currently constrained by single water levels recorded at the time of boreconstruction and this has been assumed to correspond to the end of the dry season. This wouldlikely introduce a large bias where bores were installed in periods not representative of recentconditions. •Equivalent porous media has been used to represent karstic systems. The regional equivalentporous media groundwater flow model only assumes approximate homogeneous isotropicconditions at element scales and is not suited to analysis of local karstic terrains (say, fortracking of pollutant transport). •Recharge is assumed to be diffuse; however, bypass flow via macropores and sinkholes is knownto be an important recharge mechanism. The method used to calculate recharge is empirical anddoes not include estimates of bypass flow, leading to an underestimation of recharge duringyears with above-average rainfall. •There is little understanding of actual river/aquifer interactions. This is especially the case withrespect to the flows from the groundwater system to discrete springs. •Individual springs are not considered in the Undilla1 groundwater flow model as thedistributions of the discrete pathways are too poorly understood and at a scale too small to beadequately represented. Part IV Discussion andconclusions A natural tufa damin Lawn Hill Creek, Boodjamulla National Park Photo:Russell Crosbie,CSIRO 7 Discussion This groundwater study involved several key components: (i) a literature and data review of all previous hydrogeological investigations in the catchments of the Southern Gulf rivers, that is Settlement Creek, Gregory–Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups; (ii) a regional-scale desktop data collation and analysis, including digitising data contained in hand-written and typed drilling records, evaluating groundwater levels, groundwater salinity and bore yields; (iii) a regional-scale recharge modelling and discharge mapping assessment of all aquifers in the catchments; and (iv) targeted desktop and modelling investigations of the Cambrian Limestone Aquifer (CLA) and Gilbert River Aquifer (GRA). The literature review provided good insight into all aquifers in the catchments and their current knowledge gaps. The regional-scale assessment proved useful as a screening tool for identifying spatial trends in important groundwater attributes, including the spatial extent of aquifers, ranges for groundwater levels, groundwater salinity and bore yields. It also helped identify the most promising aquifers for more detailed targeted investigations (CLA and Gilbert River Aquifer) and provided baseline datasets for use in further characterising the following properties and processes for both aquifers: (i) key hydrogeological characteristics, (ii) groundwater flow processes, (iii) water quality and chemical composition, and (iv) an initial estimate of the water balance for the CLA. 7.1 Cambrian limestone and dolostone This section summaries the conceptual model including conceptual uncertainty and water balance for the Cambrian dolostone and limestone. 7.1.1 Summary of conceptual model Hydrogeology context The Cambrian Georgina Basin aquifers underly an area of approximately 27,500 km2 within and beneath the south-west of the Southern Gulf catchments (Figure 3-2). They are dominated by limestones and dolostones (e.g. Thorntonia Limestone and Camooweal Dolostones), with maximum thicknesses of up to approximately 800 m. Mineralogical analysis of Georgina Basin rocks at stratigraphic well Carrara 1 suggests that the Camooweal Dolostone is composed of more than 80% dolomite whereas the underlying Currant Bush Limestone at this location is lithologically more diverse with variable proportions of calcite, quartz and other silicate minerals (Figure 6-4). The salinity of most groundwater samples in the CLA within the Southern Gulf catchments and adjacent areas to the west of the catchments (including the Lawn Hill Creek flow system and parts of the Southern Georgina Basin groundwater flow system) is classified as fresh to marginal, with an mean total dissolved solids (TDS) of 1096 mg/L (median = 1040 mg/L) (Figure 6-27). However, most CLA groundwaters within the Lawn Hill Creek groundwater flow system are fresh, with a trend from low salinity (from 244 mg/L) within and near the northern groundwater flow divides to 160 | Characterising groundwater resources in the Southern Gulf catchments more saline water (reaching 6112 mg/L) towards the south and south-west of the boundary of the Undilla Sub-basins (Figure 6-27). Another measure of assessing the suitability of groundwater for irrigation is the comparison of the sodium adsorption ratio (SAR) with electrical conductivity (EC) to predict the stability of soil structure in relation to irrigation water. The SAR value measures the relative concentration of sodium (Na+) to calcium (Ca2+) and magnesium (Mg2+) as calculated in ANZECC and ARMCANZ (2000). Similar to groundwater salinity, the mapped SAR data for the CLA highlight a distinctive pattern from low in the north-east of the Undilla Sub-basin (within the Southern Gulf catchments and Lawn Hill Creek groundwater flow system) to high towards the south-west of the survey area in the Southern Georgina Basin groundwater flow system (Figure 6-31). The assessment of salinity and SAR values indicates that the majority of CLA groundwater samples within the Lawn Hill Creek groundwater flow system have a hydrochemistry that is suitable for irrigation and would not affect soil structure (Figure 6-32). Recharge The Camooweal Dolostone and Thorntonia Limestone outcrop in parts of the south-western part of the Southern Gulf catchments and are covered by the Wonarah Formation or Cenozoic sediments elsewhere. Recharge rates estimated for the Camooweal Dolostone using the chloride mass balance (CMB) method range from 3.4 to 28 mm/year (with a median of 11 mm/year). For the Thorntonia Limestone, recharge rates ranged from 3.4 to 30 mm/year (with a median of 10 mm/year) (Section 6.2.1). The spatial representation of the median CMB recharge estimates (Figure 6-17) shows that there is considerable variability throughout the Georgina Basin within the Southern Gulf catchments. This is likely due to lithological variability of the aquifers; the variable thicknesses of weathering profiles and Cenozoic cover; and the occurrence of localised and preferential recharge through karst features, including cavities, sinkholes and fracture networks. Groundwater flow and residence times Previous studies assumed that groundwater contributing to the discharge at Lawn Hill Creek and Gregory River is sourced from as far west as the Alexandria–Wonarah High (Knapton et al., 2024). However, the groundwater head map constructed as part of the groundwater model development (Figure 6-15) (Section 6.4) indicates that a major regional groundwater divide separates flows to the east and flows to the south. This agrees with a recent study by Dixon-Jain et al. (2024), who named the east-flowing part of the Georgina Basin in the Southern Gulf catchments the Lawn Hill Creek flow system (Figure 6-15). There is considerable uncertainty in the dynamic range of groundwater levels in the Undilla Sub-basin used for the development of the numerical groundwater model, as the model is currently constrained by single water levels recorded at the time of bore construction. Collecting time series data at sites such as the reporting sites used in this study would reduce the uncertainty in the groundwater-level dynamics. Comprehensive hydrochemistry and environmental tracer sampling campaigns were conducted in adjacent river basins as part of the Roper River Water Resource Assessment (Taylor et al., 2024) to underpin recharge assessments and conceptual hydrogeological model development. In contrast, only ten radiocarbon analyses were conducted on groundwater samples from the Cambrian dolostone and limestone in the Georgina Basin within or in the vicinity of the Southern Gulf catchments (Figure 6-23) and no complementary tracers such as tritium or noble gases with different time ranges were analysed. Radiocarbon values ranged from less than 10 pMC (indicating the presence of old groundwater) to greater than 75 pMC (which indicates the presence of relatively young groundwater). However, the usefulness of radiocarbon data without additional complementary tracers is limited in carbonate rock aquifers because of the likely addition of ‘dead’ carbon due to either dissolution of carbonate during recharge and/or isotopic exchange during flow through the carbonate aquifers. Addition of dead carbon leads to an overestimation of calculated apparent groundwater radiocarbon ages. As demonstrated in the Roper River Water Resource Assessment groundwater hydrology activity (Taylor et al., 2024), the analysis of additional tracers such as tritium (3H) or noble gases, which cover different time ranges and which are less affected by geochemical reactions, would allow an estimate of groundwater recharge rates and residence times with higher confidence. Discharge Multiple lines of evidence were integrated in the Southern Gulf Water Resource Assessment to determine discharge processes from the Cambrian dolostone and limestone. This Assessment indicated that discharge from the CLA in the Southern Gulf catchments likely occurs via a combination of: (i) lateral outflow to streams where they are incised in the aquifer outcrop (Gregory and O’Shannassy rivers and Lawn Hill Creek); (ii) localised spring discharge; (iii) transpiration from riparian, spring-fed and other phreatophytic vegetation where the watertable for the Cambrian dolostone and limestone is shallow (i.e. ≤5 mBGL) in the south- eastern part of Lawn Hill Creek groundwater flow system; (iv) limited groundwater extraction; and (v) through geological structures and inter-basin flow. Time series data from stream gauges (Section 2.5; Figure 2-13) on the Gregory and O’Shannassy rivers indicate perennial flow (and almost perennial flow in Lawn Hill Creek), with dry-season baseflow attributed to discharge from the Thorntonia Limestone. Groundwater discharge features in the Southern Gulf catchments were also identified based on a threshold mean October actual evapotranspiration (AET), which allowed the identification of areas where dry-season AET was higher than expected and therefore potentially groundwater discharge (sections 3.9 and 5.6.4). This led to the identification of spatially extensive areas of perennial groundwater discharge in the headwaters of the Gregory and O’Shannassy rivers and Lawn Hill Creek (Figure 5-18). A two-dimensional numerical groundwater flow model was developed as part of the Southern Gulf Water Resource Assessment to examine the groundwater resources of the Undilla Sub-basin (Knapton et al., 2024), which provides baseflow to Lawn Hill Creek and Gregory River. The groundwater model successfully reproduced the observed behaviour of groundwater levels and discharge from the Cambrian dolostone and limestone aquifer in the Undilla Sub-basin. Widely spaced airborne electromagnetic (AEM) data (AusAEM with a line spacing of 20 km) are available and were used in a recent study by Dixon-Jain et al. (2024) to assess the connectivity between the Cambrian dolostone and limestone aquifers and streams and springs in the Lawn Hill flow system. This assessment suggested that inter-basin discharge from the Georgina Basin may also occur via structures that facilitate potential inter-basin connectivity from the Georgina Basin into the Constance Sandstone (a member of the South Nicholson Group in the South Nicholson Basin) and to springs. This includes culturally sensitive springs in the Boodjamulla region north of Lawn Hill, which are located along the eastern margin of the South Nicholson Basin and overlying 162 | Characterising groundwater resources in the Southern Gulf catchments Palaeozoic Georgina Basin where they meet the western edge of the Palaeoproterozoic Isa Superbasin (Dixon-Jain et al., 2024). However, the authors also emphasised that the resolution of AEM is insufficient to characterise discharge pathways with high confidence and that acquisition of higher resolution AEM data could help to better understand the relationship between subsurface geometry, geological structures and spring conceptual models. Furthermore, as already highlighted in previous sections, the lack of hydrochemistry and isotope data of springs, streamflow and aquifers within the Lawn Hill Creek flow system of the Georgina Basin has been identified as a key knowledge gap (discussed further in Section 8.2.1) and other discharge pathways to these springs such as local recharge-discharge processes within the Constance Sandstone are also plausible. 7.2 Gilbert River Formation and Rolling Downs Group This section summarises the conceptual model for the Gilbert River Formation and Rolling Downs Group including conceptual uncertainty and known components of the water balance. 7.2.1 Summary of conceptual model Hydrogeology context The geological Carpentaria Basin portion of the Great Artesian Basin (GAB) underlies most of the north-eastern parts of the Southern Gulf catchments (see Figure 2-5, Figure 2-10 and Figure 6-10). The GAB is comprised of Jurassic to Cretaceous-aged interbedded sandstones, mudstones and siltstones, which become deeper and thicker to the north of the study area (Figure 6-8 and Figure 6-9). This study is focused on the Lower Cretaceous deposits of the Gilbert River Formation andthe Lower to Upper Cretaceous Rolling Downs Group. The Rolling Downs Group comprises theNormanton, Toolebuc and Wallumbilla formations and the Allaru Mudstone. The NormantonFormation, the shallowest GAB sediment, is considered a partial aquifer, although the rest areconsidered aquitards or leaky aquitards (Table 2-1). The Gilbert River Formation comprisessandstones with minor shale lenses and hosts the Gilbert River Aquifer (GRA), the main GABaquifer in the Southern Gulf Water Resource Assessment region. Gilbert River Formation Within the Southern Gulf catchments, the Gilbert River Formation does not outcrop; it is fully covered by several hundred metres of Karumba Basin sediments and the rocks of the Rolling Downs Group, as well as Cenozoic sediments. The Gilbert River Formation is typically less than 100 m thick in the Southern Gulf catchments, but the thickness increases towards the north-west and offshore. It is typically the deepest aquifer accessed for groundwater, with bores screened at depths of 150 m, with some bores screened at over 500 mBGS. The geometry at the basin margin is not well defined due to the sparse distribution of logs with reliable stratigraphic data. These knowledge gaps mean that different representations of aquifer geometry are possible, including where the Gilbert River Formation terminates at great depths at its western extent within the Southern Gulf catchments, or where the Gilbert River Formation extends further to the basin margin where it may subcrop Cenozoic sediments. Mineralogical analysis conducted during this Assessment on cores from Dobbyn 1 exploration wells at the southern edge of the Southern Gulf catchments highlights that the Gilbert River Formation is likely to have significant permeability and is a productive aquifer at this location. Groundwater within the Gilbert River Aquifer is commonly under artesian pressure. Flow rates were initially recorded between 31 and 300 L/second, but these flows diminished substantially over time. The hydraulic conductivity is estimated to be 2 m/day (ranging from 0.1 to 10 m/day). Rolling Downs Group The Rolling Downs Group, comprised of the Normanton, Toolebuc and Wallumbilla formations, is considered an effective aquitard in the Carpentaria Basin because of the low permeability of the sediments, their lateral extent and continuity, and considerable thickness. The Wallumbilla Formation consists of marine siltstone, claystone, glauconitic sandstone and silty limestone sequences of up to 600 m thickness. The Toolebuc Formation consists of a thin unit, up to 65 m thick, of mostly limestone and carbonaceous shale. The Allaru Mudstone consists of siltstone, claystone and minor fine-grained sandstone of up to 700 m thickness. The Normanton Formation consists of sandstone and siltstone with minor glauconitic horizons of up to 300 m thickness. The Normanton Formation has partial aquifer characteristics where mean bore yields are 2 L/second. Mineralogical analysis conducted during this Assessment on cores from Dobbyn 1 exploration wells at the southern edge of the Southern Gulf catchments highlights that the Rolling Downs Group are likely to represent aquitards or leaky aquitards at this location. The Normanton Formation at the Dobbyn 1 location is unlikely to be an aquifer due to the dominance of clay minerals (e.g. smectite, kaolinite and mica) and plagioclase with less than 25% quartz (by weight). Recharge Recharge to the Gilbert River Formation and the Rolling Downs Group in the Southern Gulf catchments remains a key knowledge gap. Regional recharge estimates on Cretaceous sediments outside the Southern Gulf catchments using the CMB method are estimated to be 26 mm/year (ranging between 15 and 46 mm/year). However, the Gilbert River Formation and the Normanton Formation do not outcrop in the Southern Gulf catchments. Recharge to the Gilbert River Formation and the Rolling Downs Group is likely sourced from outside the Southern Gulf catchments and possibly leakage from contiguous sediments at the western margin of the aquifer in the Southern Gulf catchments. Environmental tracer and isotope data would be useful to investigate the source of groundwater recharge, as well as groundwater age and recharge rates, in the Southern Gulf catchments; however, these data are currently lacking in the study area. In the wider northern GAB, the volume of groundwater recharge to aquifers in the Carpentaria Basin has been estimated to be approximately 432 GL/year (Ransley et al., 2015; Smerdon et al., 2012). Groundwater flow and residence times The potentiometric surface and groundwater flow directions for the Gilbert River Formation (Ransley et al., 2015) show that flow is from the north-east boundary towards the northern coast or from the south-east outside the study area, also towards the northern coast (see Figure 6-24). While there are only a few bores whee time series measurements are available in the Gilbert River Formation, the bores that are available show that water levels are relatively stable over time (Figure 6-25) within the Southern Gulf catchments. There are insufficient bores with water-level measurements in the Rolling Downs Group to determine the potentiometric surface and groundwater flow directions in these formations. Environmental tracer and isotope data would inform the source of groundwater recharge, as well as groundwater age and recharge rates, in the Gilbert River Formation and the Rolling Downs Group in the Southern Gulf catchments, unfortunately these data are currently lacking in the study area. Two carbon isotope results in the Gilbert River Formation show that the groundwaters are relatively old with an apparent age of approximately 30,000 years, but it is likely to be much younger because high carbonate dissolution (most groundwaters are saturated with respect to dolomite and calcite) has added dead carbon to the groundwater. The lack of environmental tracer and isotope data to investigate groundwater age and recharge rates is another data gap within the GAB, as well as the larger Carpentaria Basin (Raiber et al., 2022). Discharge Considering the groundwater flow direction is towards the coast and the GAB sediments extend offshore, groundwater discharge is likely to occur offshore. However, the exact locations and volumes of discharge are unknown and are currently deemed as a knowledge gap. Moreover, insufficient information exists to quantify the water balance for the Gilbert River Formation and Rolling Downs Group in the Southern Gulf catchments. 7.3 Potential opportunities for future groundwater resource development Planning future groundwater resource developments and authorising licensed groundwater entitlements require value judgments of what is an acceptable impact on receptors such as environmental assets or existing users at a given location. These decisions can be complex and typically require considerable input from a wide range of stakeholders, particularly government regulators and communities. Scientific information to help inform these decisions includes: (i) identifying aquifers that may be potentially suitable for future groundwater resource development; (ii) characterising their depth, spatial extent, saturated thickness, hydraulic properties and water quality; (iii) conceptualising the nature of their flow systems; (iv) estimating aquifer water balances; and (v) providing initial estimates of the potential opportunities and risks for future groundwater development. Opportunities include the identification and location of hydrogeological units and the aquifers they host. Risks include the potential to changes in aquifer storage and therefore water availability to existing users or the environment (groundwater-dependent ecosystems, GDEs). The hydrogeological units of the Southern Gulf catchments (Figure 7-1) contain a variety of local-, intermediate- and regional-scale aquifers that host localised to regional-scale groundwater flow systems. The intermediate- to regional-scale CLA and Gilbert River Aquifer are present in the subsurface across large areas, collectively occurring beneath about 45% of the catchments. Given their large spatial extent, they also underlie and coincide frequently with larger areas of soil suitable for irrigated agriculture (Thomas et al., 2024). They contain mostly low-salinity water (<1000 mg/L TDS) and can yield water at a sufficient rate to support irrigation development (>10 L/second). These aquifers also store larger volumes of groundwater (gigalitres to teralitres) than local-scale aquifers. In addition, their storage and discharge characteristics are often less affected by short-term (annual) variations in recharge rates caused by inter-annual variability in rainfall. Furthermore, their larger spatial extent provides greater opportunities for groundwater resource development away from existing water users and GDEs at the land surface such as springs, spring-fed vegetation and surface water, which can be ecologically and culturally significant. In contrast, local-scale aquifers in the Southern Gulf catchments, such as fractured and weathered rock and alluvial aquifers, host local-scale groundwater systems that are highly variable in composition, salinity and yield. They also have a small and variable spatial extent and lower storage compared to the larger aquifers. This limits groundwater resource development to localised opportunities such as stock and domestic use, or as a conjunctive water resource (i.e. combined use of surface water, groundwater or rainwater). The Assessment identified five hydrogeological units hosting aquifers that may have potential for future groundwater resource development in the Southern Gulf catchments (Table 7-1): •Cambrian limestone and dolostone •Cretaceous rocks •Cenozoic alluvium •Proterozoic igneous rocks •Proterozoic metasedimentary and metamorphic rocks. Table 7-1 Opportunity-level estimates of the potential scale of groundwater resource development opportunities in the Southern Gulf catchments For locations of the hydrogeological units see Figure 7-1. The indicative scale of the groundwater resource is based on the magnitude of the inputs and outputs of the groundwater balance. The actual scale will depend upon government and community acceptance of potential impacts of groundwater-dependent ecosystems and existing groundwater users. For more information on this table please contact CSIRO on enquiries@csiro.au For more information on this table please contact CSIRO on enquiries@csiro.au A map of a country Description automatically generated Figure 7-1 Hydrogeological units with potential for future groundwater resource development Presents the spatial extent of the outcropping and subcropping component of each hydrogeological unit with the majority of the Cenozoic cover removed (except the alluvium). Figure adapted from Fig SW-2 in CSIRO (2019a). 7.3.1 Cambrian Limestone Aquifer While the current level of knowledge for the CLA hosted in the Cambrian limestone and dolostone is limited in terms of pre-feasibility information, the following information indicates it offers some potential for future development: (i) the moderate spatial extent of the outcropping/subcropping area in the catchments (12,400 km2, see Figure 7-1) coincides with areas of cracking clay soils potentially suitable for agricultural intensification (Thomas et al., 2024); (ii) the aquifers can be intersected by drilling at relatively shallow depths in the outcropping and subcropping areas (mostly <100 mBGL) (see Section 6.1.1); (iii) moderately yielding bores (ranging up to 20 L/second) have the potential to support groundwater-based irrigation, however, further aquifer testing is required (see Section 5.4); (iv) the depth to pump groundwater to the surface is less than 75 mBGL across most areas of the CLA (see Section 6.2.1), though groundwater-level data are sparse, and larger areas of shallow depths to groundwater (<10 mBGL) are likely to occur around the middle to lower reaches of the prescribed watercourses of the Gregory River and Lawn Hill Creek where the aquifer discharges; and (v) hosts good-quality water suitable for a variety of irrigated crops (mostly <1000 mg/L TDS). Initial modelling of the mean annual groundwater balance for the CLA in the south-western Southern Gulf catchments (see Section 6.4), suggests potentially promising opportunities for future groundwater resource development based on the magnitude of groundwater flows. However, the initial model has limitations due to the data-sparse nature of the CLA in this area including a lack of: (i) spatial and temporal groundwater-level data, (ii) gauging of spring flow at discrete springs, and (iii) estimates of mean annual recharge using a variety of independent methods. In addition, recent modelling of the CLA by Knapton et al. (2023) has highlighted the potential sensitivity of the CLA groundwater balance to climate variables. In this study, these factors have been accounted for when considering an indicative scale of the potential resource. Applying both the 5th percentile mean annual recharge rate of 3 mm/year (Section 5.5.3) and the lowest modelled recharge rate of 7 mm/year (see Section 6.4) to the portion of the CLA in the south-west of the catchments (13,246 km2) yields a potential annual recharge flux of between 40 and 93 GL/year. Assuming 20% of this range in annual recharge fluxes may be available for potential future groundwater resource development, this would equate to volumes ranging from 8 to 19 GL/year. However, this does not account for groundwater inflows from catchments upgradient of the CLA outside the south-western parts of the Southern Gulf catchments. Therefore, the indicative scale of the groundwater resource in the CLA is estimated to be between 10 and 20 GL/year (Table 7-1). However, this requires further investigation. 7.3.2 Gilbert River Aquifer and Rolling Downs Group Although the current level of knowledge for the Gilbert River Aquifer (GRA) hosted in the Late Jurassic to Cretaceous Gilbert River Formation is limited in terms of pre-feasibility information, the following information indicates its potential for future development: (i) the large spatial extent of the aquifer in the subsurface (about 44,600 km2, or about 40% of the catchments, see Figure 7-1) that coincides with soils potentially suitable for agricultural intensification (Thomas et al., 2024); (ii)the aquifer can be intersected by drilling at depths of mostly less than 400 mBGL (see Section6.1.2) along the south-western margin of the Gilbert River Aquifer in the north-east of theSouthern Gulf catchments, an area where groundwater is mostly fresh and less than 1000 mg/L TDS (see Section 5.3.1); and (iii) the aquifer is either artesian or close to artesian thereby reducing the cost of pumping the groundwater to the surface (see Section 6.2.2), though groundwater-level data are sparse. In bores towards the Gulf of Carpentaria, the higher TDS means that only moderately tolerant and tolerant crops (ANZECC and ARMCANZ, 2000) could be grown using water from the Gilbert River Aquifer Formation or Rolling Downs Group (see Figure 6-43 and Figure 6-44). Moreover, many of the Gilbert River Aquifer Formation or Rolling Downs Group bores are oversaturated with respect to calcite and dolomite, so scaling of irrigation equipment and soils could potentially be a problem. At this point in time, insufficient information exists to quantify the water balance of the Gilbert River Formation in the Southern Gulf catchments. Furthermore, the aquifers: (i) have sparse temporal water-level information, (ii) dip steeply in the subsurface indicating that they shift across different areas from semi-confined to confined conditions, and (iii) support active water licences for community water supplies at Burketown and Gununa, as well potentially supporting ecologically and culturally important flora and fauna in the Gulf of Carpentaria’s marine environment via coastal and submarine groundwater discharge. In the absence of further hydrogeological investigations (drilling and pump testing) and hydrological risk assessment modelling to evaluate groundwater extraction impacts on existing water users and GDEs, a conservative value of less than, or equal to 5 GL/year has been assumed as an indicative scale of the resource for potential future groundwater resource development (Table 7-1). However, this requires further investigation. 8 Summary and conclusions 8.1 Opportunities and constraints for future groundwater development The literature review and regional desktop assessment of the existing hydrogeological information served as an effective screening tool for identifying potential groundwater resource development opportunities within the Southern Gulf catchments. Insights gained from the desktop analyses and modelling have improved the understanding of both the Cambrian Limestone Aquifer (CLA) and the Jurassic to Cretaceous Gilbert River Aquifer (GRA, hosted within the Gilber River Formation), which offer the most promising opportunities for future development. The greatest opportunities for the CLA exist along the south-western part of the Southern Gulf catchments boundary in the aquifer outcrop. Opportunities are limited where the aquifer co- occurs with the mid- to lower reaches of the prescribed watercourses. This is where groundwater discharge from the CLA supports ecologically and culturally important springs and seeps that support streamflow in the Gregory River and Lawn Hill Creek in the NGMA of the Gregory River Subcatchment Area of the Gulf Water Plan. The greatest opportunities for the GRA exist along the south-western margin of the Jurassic to Cretaceous rocks where the formation is less than 500 m deep, and where the groundwater is fresh. Opportunities are limited near ecologically and culturally important springs and seeps, as well as near existing licensed water users (such as the communities of Burketown and Gununa), and where the GRA is prohibitively deep (>500 mBGL) and/or where the groundwater is brackish (>2000 mg/L TDS). Opportunities for small-scale (<0.5 GL/year) localised developments exist within the Cenozoic alluvium, Proterozoic igneous rocks, and Proterozoic metasedimentary and metamorphic rocks. Within the alluvial aquifers in the Nicholson, Gregory and Leichhardt rivers, and Settlement Creek and its tributaries, there may be opportunities for multiple small-scale (<0.5 GL/year) localised developments or as a conjunctive water resource where surface water is available. Opportunities are likely to be limited where the saturated thickness of the aquifer is thin (<10 m), as well as in areas within 1 km of the prescribed watercourses of the Gregory and Nicholson rivers, whose streamflow is supported by groundwater discharge from the alluvium. Proterozoic igneous rocks only have the potential for small-scale (<0.5 GL/year) localised developments (i.e. mostly suited to stock and domestic water supplies) where fracturing and weathering is high. Bore yields can exceed 20 L/second where the density and interconnectivity of secondary porosity features (fractures, faults and joints) is high, though in most cases aquifer storage is limited. Proterozoic metasedimentary and metamorphic rocks likely only have the potential for small-scale (<0.25 GL/year) localised developments (i.e. mostly suited to stock and domestic water supplies) in the outcropping area where fracturing and weathering is high. 8.2 Potential options for future work The regional desktop assessment of all available hydrogeological data proved useful as a screening tool for identifying potential groundwater resource development opportunities across the catchments of the Southern Gulf rivers, that is Settlement Creek, Gregory–Nicholson River and Leichhardt River, the Morning Inlet catchments and the Wellesley island groups. However, multiple knowledge gaps and opportunities to close these gaps were identified, with many aligning with the original proposed targeted fieldwork (Section 3.1). 8.2.1 Cambrian Limestone Aquifer Multiple knowledge gaps and opportunities for further work were identified for the CLA. These include: 1. Water-level data: as part of the groundwater numerical model development for the Southern Gulf Water Resource Assessment, Knapton et al. (2024) highlighted that water-level time series data could reduce the uncertainty in the groundwater-level dynamics for the Cambrian dolostone and limestone aquifers and their interaction with surface water features. 2. Recharge rates: groundwater chloride concentration values within the Southern Gulf catchments are sparse for some surface geology units. Additional measurements would reduce their uncertainty. Furthermore, there are no rainfall chemistry and isotope observation points within the Southern Gulf catchments, and rainfall chemistry measurements could reduce the uncertainty of chloride mass balance recharge estimates. 3. Residence times tracers: a lack of environmental tracers has been identified in this, and in previous studies as key knowledge gaps in the assessment of recharge processes and surface water – groundwater connectivity. This includes the simultaneous application of multiple residence/age tracers such as 3H, 14C and noble gases along inferred groundwater flow paths (similar to the example shown in Figure 8-1 for the Gilbert River Aquifer and as demonstrated by Taylor et al. (2024) in the adjacent Roper River Water Resource Assessment). This would allow the identification of recharge processes including preferential recharge through karst features and the independent quantification of recharge processes. 4. Isotopic fingerprinting of springs and groundwaters to determine spring source aquifer: developing a hydrochemical and isotopic fingerprinting framework (similar to the framework developed by Raiber et al. (2024) for the south-eastern part of the Great Artesian Basin, GAB) could help to determine sources of discharge to Lawn Hill Creek (Lawn Hill Gorge) and Gregory River. A diagram of a geological formation Description automatically generated with medium confidence Figure 8-1 Example of isotope tracer sampling opportunities for the Gilbert River Aquifer in the Southern Gulf catchments The graph gives examples of approximate age scales and tracer concentrations expected at different flow path intervals, with expected timescales ranging from millennia to several hundred thousand or greater than 1 million years. Modified from Raiber et al. (2022) and Suckow (2014). 8.2.2 Gilbert River Aquifer and Rolling Downs Group The GRA and Rolling Downs Group within the Carpentaria Basin have been the subject of several hydrogeological and hydrochemical investigations over the past two decades. Nevertheless, knowledge gaps remain, and these could be addressed by: • improving the mapping of the Carpentaria Basin geometry at the basin margin • performing mineralogical characterisation (X-ray diffraction (XRD) and X-ray fluorescence (XRF)) on additional cores to verify the thickness and location of transmissive zones, particularly in the Rolling Downs Group • determining the location and source of recharge to the GRA and Rolling Downs Group, which is likely from outside the Southern Gulf catchments as the sediments do not outcrop in the area or through overlying strata at the basin margin • estimating the location and volume of groundwater discharge offshore to help quantify the water balance for the GRA and Rolling Downs Group in the Southern Gulf catchments • undertaking multi-tracer sampling (e.g. 3H, 14C, 36Cl and stable and radioactive noble gases) along inferred flow paths to determine residence times in the Gilbert River Formation and the Rolling Downs Group (Figure 8-1), and obtain water chemistry values for areas currently lacking data • obtaining more spatial and temporal hydraulic head data for the GRA and Rolling Downs Group, including the installation of water-level loggers in sub-artesian bores • undertaking pumping tests to help constrain the water balance of the GRA and Rolling Downs Group. 8.2.3 Alluvial aquifers The alluvial aquifers within the Southern Gulf catchments are poorly characterised due to a lack of groundwater bores and monitoring bores with reliable lithological, water chemistry and hydraulic data. As a result, many knowledge gaps remain, and these could be addressed by: 1. Geophysical surveys: geophysical surveys such as ground-based transient electromagnetic surveys or airborne electromagnetic (AEM) surveys could provide insights into the extent, thickness and internal architecture of alluvial aquifers and identify and localise the connectivity with streams. 2. Targeted drilling and instrumentation of bores: there are currently only very few groundwater bores with reliable lithological, water chemistry and hydraulic data; drilling of additional bores and instrumentation of bores with data loggers would support the understanding of watertable fluctuations, aquifer water quality and yields and the connectivity of aquifers with surface waters. 3. Sampling of tracers and hydrochemistry: the collection and analysis of alluvial groundwaters can help to understand residence times, groundwater – surface water connectivity and alluvial recharge processes. 4. Pumping tests: pumping tests on alluvial bores would help to understand bore yields and the suitability of alluvial aquifers for groundwater extraction. References Alcalá FJ, Cantón Y, Contreras S, Were A, Serrano-Ortiz P, Puigdefábregas J, Solé-Benet A, Custodio E and Domingo F (2011) Diffuse and concentrated recharge evaluation using physical and tracer techniques: results from a semiarid carbonate massif aquifer in southeastern Spain. Environmental Earth Sciences 62(3), 541–557. DOI: 10.1007/s12665-010-0546-y. Anderson VG (1945) Some effects of atmospheric evaporation and transpiration on the composition of natural water in Australia (continued). 4. Underground waters in riverless areas. J. Aust. Chem. Inst 12, 83–98. Anderson J, Carson C, Lewis C, Cross A, Carr L, Henson P and Kositcin N (2019) Exploring for the future: new U-Pb geochronology for the South Nicholson region and implications for stratigraphic correlations. ASEG Extended Abstracts, 2019(1), 1–5. https://doi.org/10.1080/22020586.2019.12073005. ANZECC and ARMCANZ (2000) Australian and New Zealand guidelines for fresh and marine water quality. Volume 1. Australian and New Zealand Governments and Australian state and territory governments, Canberra ACT, Australia. Available online: http://www.waterquality.gov.au/anz-guidelines/. Australian Bureau of Statistics (2021) Australian Bureau of Statistics, 2021 Census data. Australia’s national statistical agency, Australian Government. AWRC (1976) Review of Australia’s Water Resources 1975, Australian Water Resources Council, Department of National Resources, Australian Government Publishing Service, Canberra. Bardwell N and Grey D (2016) Hydrogeochemistry of Queensland: Data Release: Accompanying Notes. CSIRO, Australia. EP156406 22p. GUID: B419D35B-A8D5-47B8-9FB6-85FB4D98D1CA https://repo.bioregionalassessments.gov.au/metadata/B419D35B-A8D5-47B8-9FB6- 85FB4D98D1CA. Bailey AHE, Bradshaw BE, Palu TJ, Wang L, Jarrett AJM, Orr ML, Lech ME, Evenden C, Arnold D, Reese B, Skeers N, Woods M, Dehelean A, Lawson C and Hall LS (2020) Shale gas prospectivity of the Isa GBA region. Technical appendix for the Geological and Bioregional Assessment: Stage 2. Department of the Environment and Energy, Bureau of Meteorology, CSIRO and Geoscience Australia, Australia. Bazuhair AS and Wood WW (1996) Chloride Mass-Balance Method for Estimating Ground Water Recharge in Arid Areas - Examples From Western Saudi Arabia. Journal of Hydrology 186(1– 4), 153–159. BoM (2012) Australian hydrological geospatial fabric (Geofabric): Product guide Version 2.1. Australian Government, Bureau of Meteorology. BoM (2019) Normalized Difference Vegetation Index (NDVI): AVHRR, without atmospheric correction, Australia coverage. http://www.auscover.org.au/purl/avhrr-ndvi-bom. Accessed April 2019. Bradshaw BE, Spencer LK, Lahtinen AC, Khider K, Ryan DJ, Colwell JB, Chirinos A and Bradshaw J (2009) Queensland carbon dioxide geological storage atlas. Queensland Department of Employment, Economic Development and Innovation, Brisbane. Bradshaw BE, Orr ML, Bailey AHE, Palu TJ and Hall LS (2018) Northern Lawn Hill Platform depth structure and isochore mapping update. Record 2018/47. Geoscience Australia, Canberra. Viewed 08 March 2019, http://dx.doi.org/10.11636/Record.2018.047. Brodie RS, Kilgour B, Jacobson G and Lau JE (2019) Hydrogeology Map of Australia (G.Jacobson and JE.Lau Hydrogeology Map). Geoscience Australia, Canberra., https://ecat.ga.gov.au/geonetwork/srv/eng/catalog.search#/metadata/32368. Buchanan S, Dixon-Jain P, Martinez J, Raiber M, Kumar PR, Woods M, Arnold D, A. D and Skeers N (2020) Hydrogeology and groundwater systems of the Isa GBA region. Technical appendix for the Geological and Bioregional Assessment: Stage 2. Department of the Environment and Energy, Bureau of Meteorology, CSIRO and Geoscience Australia, Australia., https://www.bioregionalassessments.gov.au/sites/default/files/isa_stage_2_- _technical_appendix_hydrogeology.pdf. Budget Strategy and Outlook (2024) Budget 2024-25. A future made in Australia. Viewed 11 September 2024, https://budget.gov.au/content/factsheets/download/factsheet-fmia.pdf. Budyko MI (1974) Climate and life. Academic Press. Available online: https://www.sciencedirect.com/science/article/pii/S0074614209600012. Carr L, Ayling B, Morris R, Higgins K, Edwards D, Khider K, Smith T, Lawrie J, Morse M, Boreham C, Tenthorey E, Buckler T, Cox W, Hatch L, Woods M, Cortese A, Southby C and Chirinos A (2016) Georgina Basin Geoscience Data Package. Australia G, Canberra, Australia. Carr LK, Southby C, Henson P, Costelloe R, Anderson JR, Jarrett AJM, Carson CJ, MacFarlane SK, Gorton J, Hutton L, Troup A, Williams B, Khider K, Bailey AHE and Fomin T (2019) Exploring for the Future: South Nicholson Basin Geological summary and seismic data interpretation. Record 2019/21. Geoscience Australia, Canberra. http://dx.doi.org/10.11636/Record.2019.021. Cartwright I, Weaver TR, Stone D and Reid M (2007) Constraining modern and historical recharge from bore hydrographs, H-3, C-14 and chloride concentrations: Applications to dual-porosity aquifers in dryland salinity areas, Murray Basin, Australia. Journal of Hydrology 332(1–2), 69– 92. Castellazzi P, Gao S, Pritchard J, Ponce Reyes R and Stratford D (2024) Detecting springs and groundwater-dependent vegetation in data-scarce regions of Australia combining citizen science, GRACE, and optical/radar imagery, Remote Sensing of Environment. Available at SSRN: https://ssrn.com/abstract=4827285 or http://dx.doi.org/10.2139/ssrn.4827285. Cook AG and Jell JS (2013) Paleogene and Neogene. Geology of Queensland. Geological Survey of Queensland, Brisbane. In: Jell PA (ed.). Cook AG, S.E. B and Draper JJ (2013) Chapter 7: Post-orogenic Mesozoic basins and magmatism. In: Jell PA (ed.), Geology of Queensland. Geological Survey of Queensland, Brisbane. Cook PG, Walker GR and Jolly ID (1989) Spatial Variability of Groundwater Recharge in a Semiarid Region. Journal of Hydrology 111, 195–212. Coxon G, Freer J, Westerberg IK, Wagener T, Woods R and Smith PJ (2015) A novel framework for discharge uncertainty quantification applied to 500 UK gauging stations. Water Resources Research 51(7), 5531–5546. DOI: https://doi.org/10.1002/2014WR016532. Craig H (1961) Isotopic variations in Meteoric Waters. Science, 1702-1703. Crosbie R, Pickett T, Mpelasoka F, Hodgson G, Charles S and Barron O (2013) An assessment of the climate change impacts on groundwater recharge at a continental scale using a probabilistic approach with an ensemble of GCMs. Climatic Change 117, 41–53. DOI: 10.1007/s10584- 012-0558-6. Crosbie R, Raiber M, Cui T and Viney N (2015) Blending field observations and AWRA outputs for groundwater recharge estimation in a coal basin in eastern Australia. In: MODSIM2015, 21st International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, December 2015, 2033–2039. http://www.mssanz.org.au/modsim2015/L2032/crosbie.pdf. Crosbie RS, Jolly ID, Leaney FW and Petheram C (2010) Can the dataset of field based recharge estimates in Australia be used to predict recharge in data-poor areas? Hydrology and Earth System Sciences 14(10), 2023–2038. DOI: 10.5194/hess-14-2023-2010. Crosbie RS, Peeters LJM, Herron N, McVicar TR and Herr A (2018) Estimating groundwater recharge and its associated uncertainty: Use of regression kriging and the chloride mass balance method. Journal of Hydrology 561, 1063–1080. DOI: https://doi.org/10.1016/j.jhydrol.2017.08.003. Crosbie RS and Rachakonda PK (2021) Constraining probabilistic chloride mass-balance recharge estimates using baseflow and remotely sensed evapotranspiration: the Cambrian Limestone Aquifer in northern Australia. Hydrogeology Journal 29, 1399–1419. DOI: 10.1007/s10040- 021-02323-1. Crosbie RS, Raiber M, Wilkins A, Dawes W, Louth-Robins T and Gao L (2022) Quantifying diffuse recharge to groundwater systems of the Great Artesian Basin, the NSW coalfields and surrounds. CSIRO Report EP2022–2355. CSIRO (2009a) Water in the Gulf of Carpentaria Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. xl + 479pp. Available online: https://www.clw.csiro.au/publications/waterforahealthycountry/nasy/documents/GulfOfCarpentaria/NASY-Gulf-of-Carpentaria-Drainage-Division-Report.pdf. CSIRO (2009b) Water in the Roper Region of the Gulf of Carpentaria Drainage Division: a report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO, Canberra, ACT. CSIRO (2009c) Water in the South-West Gulf region, pp 121–185 in CSIRO (2009) Water in the Gulf of Carpentaria Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. xl + 479pp. Available online: https://www.clw.csiro.au/publications/waterforahealthycountry/nasy/documents/GulfOfCarpentaria/NASY-South-West-Gulf-Region-Report.pdf. Department of Environment and Natural Resources (NT) (2019) Springs of the Northern Territory. [spatial] Geoscience A, Canberra, Viewed 20 March 2019, http://www.ntlis.nt.gov.au/metadata/export_data?type=html&metadata_id=E9D78F668FC02B33E040CD9B214403C3. Department of Environment Parks and Water Security (2000a) Melaleuca Survey of the Northern Territory. ANZLIC Identifier: 2DBCB7711FE606B6E040CD9B0F274EFE, http://www.ntlis.nt.gov.au/metadata/export_data?metadata_id=2DBCB7711FE606B6E040CD9B0F274EFE&type=html. Department of Environment, Parks and Water Security (2000b) Bores sites in the NorthernTerritory. ANZLIC Identifier: 2DBCB7711FC106B6E040CD9B0F274EFE, data exported 12/07/2023, http://www.ntlis.nt.gov.au/metadata/export_data?type=html&metadata_id=2DBCB7711FC106B6E040CD9B0F274EFE. Department of Environment, Parks and Water Security (2014) Digital groundwater database for the Northern Territory, supplied by the Department of Environment, Parks and Water Security, Copyright – Northern Territory of Australia. Data downloaded from the Northern Territory Government (NTG) Open Data Portal in September 2021. https://data.nt.gov.au/dataset/nt-bore-locations-water-quality-and-groundwater-levels. Department of Environment, Parks and Water Security, NT Government (2022) Land use mapping project of the Northern Territory (2016–2022) (LUMP). Open data portal, NT Government. Sourced 28 March 2023, https://data.nt.gov.au/dataset/land-use-mapping-project-of-the- northern-territory-2016-current-lump. Department of Environment, Parks and Water Security (2021) NR maps, the Department of Environment and Natural Resources web mapping tool for natural and cultural research data for the Northern Territory. Data accessed January 2021., https://nrmaps.nt.gov.au/nrmaps.html. Department of Environment, Science and Innovation (2021) Springs database. https://www.data.qld.gov.au/dataset/springs#:~:text=The%20Queensland%20Springs%20Database%20provides%20a%20comprehensive%20catalogue,and%20any%20associated%20surface%20expression%20groundwater%20dependent%20ecosystems. Department of Regional Development Manufacturing and Water (2021) Spatial dataset displaying the location of water licenses and permits across Queensland. Queensland Department of Regional Development, Manufacturing and Water. © State of Queensland 2021. Department of Regional Development, Manufacturing and Water (2023) Groundwater Database – Queensland. https://www.data.qld.gov.au/dataset/groundwater-database-queensland. Department of State Development, Manufacturing, Infrastructure and Planning (2019) North west Queensland economic diversification strategy 2019. Viewed 12 September 2024, https://www.statedevelopment.qld.gov.au/regions/regional-priorities/a-strong-and- prosperous-north-west-queensland/north-west-queensland-economic-diversification- strategy. Diersch H-JG (2008) FEFLOW® 5.4 Users Manual. WASY – Institute for Water Resources Planning and System Research, Berlin, Germany. Dixon-Jain P, Bishop C, Lester J, Orlov C, McPherson A, Pho G, Flower C, Kilgour P, Lawson S, Vizy J, and Lewis S (2024) Hydrogeology and groundwater systems of the South Nicholson and Georgina basins, Northern Territory and Queensland. Record 2024/37. Geoscience Australia, Canberra. https://dx.doi.org/10.26186/149730. Doherty J (2003) Ground water model calibration using pilot points and regularization. Groundwater 41(2), 170–177. DOI: 10.1111/j.1745-6584.2003.tb02580.x. Doherty J (2024) PEST_HP: a parameter estimation software package for highly parameterized models. Watermark Numerical Computing, Brisbane, Australia. Doody TM, Barron OV, Dowsley K, Emelyanova I, Fawcett J, Overton IC, Pritchard JL, Van Dijk AIJM and Warren G (2017) Continental mapping of groundwater dependent ecosystems: A methodological framework to integrate diverse data and expert opinion. Journal of Hydrology: Regional Studies 10, 61–81. DOI: https://doi.org/10.1016/j.ejrh.2017.01.003. DOR (2021a) Groundwater Database - Queensland. Database storing registered water bore data from private water bores and Queensland Government groundwater investigation and monitoring bores. © State of Queensland (Department of Resources ) 2021. Licensed under Single Supply, licence sourced in September 2021. http://qldspatial.information.qld.gov.au/catalogue/custom/search.page?q=%22Groundwater%20Database%20-%20Queensland%22. DOR (2021b) Queensland Globe is an online interactive tool that displays physical, geographical and spatial data about a particular location into map format. Queensland Government, Department of Resources, Queensland Globe. https://qldglobe.information.qld.gov.au/. Drysdale RN, Taylor MP and Ihlenfeld C (2002) Factors controlling the chemical evolution of travertine-depositing rivers of the Barkly karst, northern Australia. Hydrological Processes 16(15), 2941–2962. DOI: 10.1002/hyp.1078. Efron B and Tibshirani RJ (1994) An introduction to the bootstrap. CRC press. Environment Australia (2001) A Directory of Important Wetlands in Australia, Third Edition., Environment Australia, Canberra. Eriksson E (1985) Principles and applications of hydrochemistry. Chapman and Hall Ltd, New York. Foley EK, Roberts EM and EM Knutsen (2022) Deciphering Late Cretaceous palaeo-river catchments in eastern Australia: recognition of distinct northern and southern drainage basins. Basin Research, 34 (2). pp. 590-617. Gallant J, Wilson N, Dowling T, Read A and Inskeep C (2011) SRTM-derived 1 second digital elevation models version 1.0. Geoscience Australia: Canberra, ACT. DOI: http://pid.geoscience.gov.au/dataset/ga/72759. Geognostics (2021) OZ SEEBASE® 2021 (March 2021). Geognostics Australia Pty Ltd, https://www.geognostics.com/oz-seebase-2021. Geoscience Australia (2021) DEA Fractional Cover Percentiles (Landsat). Geoscience Australia. Available online: https://cmi.ga.gov.au/data-products/dea/630/dea-fractional-cover- percentiles-landsat. Geoscience Australia and Australian Stratigraphy Commission (2021) Australian Stratigraphic Units Database. https://www.ga.gov.au/data-pubs/datastandards/stratigraphic-units. Geoscience Australia Portal (2023) https://portal.ga.gov.au/ . Geoscience Australia (2024). South Nicholson National Drilling Initiative. https://www.eftf.ga.gov.au/south-nicholson-national-drilling-initiative. Glenn EP, Doody TM, Guerschman JP, Huete AR, King EA, McVicar TR, Van Dijk AIJM, Van Niel TG, Yebra M and Zhang Y (2011) Actual evapotranspiration estimation by ground and remote sensing methods: the Australian experience. Hydrological Processes 25(26), 4103–4116. DOI: 10.1002/hyp.8391. Grayson R, Argent R, Nathan R, McMahon T and Mein R (1996) Hydrological recipes: Estimation techniques in Australian hydrology. CRC Catchment Hydrology, Melbourne. Grimes KG (1974) Mesozoic and Cainozoic geology of the Lawn Hill, Westmoreland, Mornington and Cape Van Diemen 1:1250 000 Sheet area, Queensland. Record 1974/06. Geoscience Australia, Canberra. Grundy MJ, Rossel RAV, Searle RD, Wilson PL, Chen C and Gregory LJ (2015) Soil and Landscape Grid of Australia. Soil Research 53(8), 835–844. DOI: http://dx.doi.org/10.1071/SR15191. Guerschman JP, McVicar TR, Vleeshower J, Van Niel TG, Peña-Arancibia JL and Chen Y (2022) Estimating actual evapotranspiration at field-to-continent scales by calibrating the CMRSET algorithm with MODIS, VIIRS, Landsat and Sentinel-2 data. Journal of Hydrology 605, 127318. DOI: https://doi.org/10.1016/j.jhydrol.2021.127318. Guerschman JP, Warren G, Byrne G, Lymburner L, Mueller N and Van Dijk AI (2011) MODIS-based standing water detection for flood and large reservoir mapping: algorithm development and applications for the Australian continent. CSIRO. Güler C, Thyne GD, McCray JE, Turner, KA (2002) Evaluation of graphical and multivariate statistical methods for classification of water chemistry data. Hydrogeol. J. 10: 455–474. doi:10.1007/s10040-002-0196-6. Hengl T, Heuvelink GBM and Stein A (2004) A generic framework for spatial prediction of soil variables based on regression-kriging. Geoderma 120(1–2), 75–93. DOI: http://dx.doi.org/10.1016/j.geoderma.2003.08.018. Herbert GC (2000) Hydrogeology of the Karumba Basin western Cape York Peninsula. Post graduate diploma project, School of Natural Resource Sciences, Brisbane. Herczeg AL, Torgersen T, Chivas AR and Habermehl MA (1991) Geochemistry of ground waters from the Great Artesian Basin, Australia. Journal of Hydrology 126(3–4), 225–245. DOI: 10.1016/0022-1694(91)90158-E. Hermoso V, Ward DP and Kennard MJ (2013) Prioritizing refugia for freshwater biodiversity conservation in highly seasonal ecosystems. Diversity and Distributions 19(8), 1031–1042. DOI: 10.1111/ddi.12082. Hollins SE, Hughes CE, Jagoda C, Dioni I, Cendon K, Karina T and Meridith (2018) Rainfall isotope variations over the Australian continent - implicationsfor hydrology and isotope applications. Science of the Total Environment 645, 630-645. Horn AM, Derrington EA, Herbert GC, Lait RW and Hillier JR (1995) Groundwater Resources of Cape York Peninsula. (Cape York Peninsula Land Use Strategy, Office of the Co-ordinator General of Queensland, Brisbane, Department of the Environment, Sport and Territories, Canberra, Queensland Department of Primary Industries, Brisbane and Mareeba, and Australian Geological Survey Organisation, Mareeba). Howcroft W, Cartwright I, Fifield LK and Cendón D (2017) Differences in groundwater and chloride residence times in saline groundwater: The Barwon River Catchment of Southeast Australia. Chemical Geology 451, 154–168. Ingram JA (1972) Donors Hill, Queensland Sheet SE/54-10. 1:250 000 Geological Series Explanatory Notes. Jeffrey SJ, Carter JO, Moodie KB and Beswick AR (2001) Using spatial interpolation to construct a comprehensive archive of Australian climate data. Environmental Modelling and Software 16, 309-330. DOI: 10.1016/S1364-8152(01)00008-1. Jolly P and Tickell S (2011) Tufa Dams of Northern Australia. NT Branch of IAH. Jones DA, Wang W and Fawcett R (2009) High-quality spatial climate data-sets for Australia. Australian Meteorological and Oceanographic Journal 58(4), 233–248. Kalma JD, McVicar TR and McCabe MF (2008) Estimating Land Surface Evaporation: A Review of Methods Using Remotely Sensed Surface Temperature Data. Surveys in Geophysics 29(4–5), 421–469. DOI: 10.1007/s10712-008-9037-z. Kim JH and Jackson RB (2012) A Global Analysis of Groundwater Recharge for Vegetation, Climate, and Soils. Vadose Zone Journal 11(1). DOI: 10.2136/vzj2011.0021RA. King E, van Niel T, van Dijk A, Wang Z, Paget M, Raupach T, Guerschman J, Haverd V, McVicar T, Miltenberg I, Raupach M, Renzullo L and Zhang Y (2011) Actual Evapotranspiration Estimates for Australia Intercomparison and Evaluation. CSIRO: Water for a Healthy Country National Research Flagship, Canberra. Klohn Clippen Berger (2016) Hydrogeological assessment of the Great Artesian Basin: Characterisation of aquifer groups, Carpentaria Basin. Department of Natural Resources and Mines (Queensland), Brisbane. Knapton A, Taylor AR and Crosbie RS (2024) Groundwater flow modelling of the Cambrian Limestone Aquifer in the Undilla Sub-basin, Georgina Basin using a finite element groundwater flow model. A technical report from the CSIRO Southern Gulf Water Resource Assessment for the National Water Grid. CSIRO, Australia. Krause CE, Newey V, Alger MJ and Lymburner L (2021) Mapping and Monitoring the Multi-Decadal Dynamics of Australia’s Open Waterbodies Using Landsat. Remote Sensing 13(8), 1437. Kruse P and Dunster J (2013) Chapter 28: Georgina Basin: in Ahmad M and Munson TJ (compilers) ‘Geology and Mineral Resources of the Northern Territory’. Northern Territory Geological Survey, Special Publication 5. Lewis A, Oliver S, Lymburner L, Evans B, Wyborn L, Mueller N, Raevksi G, Hooke J, Woodcock R, Sixsmith J, Wu W, Tan P, Li F, Killough B, Minchin S, Roberts D, Ayers D, Bala B, Dwyer J, Dekker A, Dhu T, Hicks A, Ip A, Purss M, Richards C, Sagar S, Trenham C, Wang P and Wang L- W (2017) The Australian Geoscience Data Cube — Foundations and lessons learned. Remote Sensing of Environment 202, 276–292. DOI: https://doi.org/10.1016/j.rse.2017.03.015. Ley-Cooper Y and Brodie R (2018) AusAEM Year 1 NT/QLD Airborne Electromagnetic Survey; GA Layered Earth Inversion Products [data set], Geoscience Australia, Canberra, https://pid.geoscience.gov.au/dataset/ga/132709. Lynne V and Hollick M (1979) Stochastic time variable rainfall runoff modelling. In: Hydrology and Water Resources Symposium, 89–92. Martinez J, Raiber M and Cox ME (2015) Assessment of groundwater-surface interaction via combination of long-term hydrochemical dataset and isotope hydrology: an example from the headwaters of the Condamine River, Southeast Queensland, Australia. Science of the Total Environment, 536, 499-516. Matthews I (1992) Alpurrurulam Groundwater Source and Borefield Review. Power and Water Authority. Technical Report WRA92008. Available online: https://territorystories.nt.gov.au/10070/674779/0/0. McConachie B, Dunster J, Wellman P, Denaro T, Pain C, Habermehl M and Draper J (1997) Carpentaria Lowlands and Gulf of Carpentaria regions. In: Bain J and Draper J (eds), North Queensland Geology. AGSO Bulletin 240. Queensland Geology 9. Australian Geological Survey Organisation and Department of Mines and Energy Queensland, Canberra. McEniery M (1980) Groundwater resources - north Queensland. In: Henderson R and Stephenson P (eds), The geology and geophysics of northeastern Australia Volume 1. Geological Society of Australia, Queensland, Brisbane. McJannet D, Marvanek S, Kinsey-Henderson A, Petheram C and Wallace J (2014) Persistence of in- stream waterholes in ephemeral rivers of tropical northern Australia and potential impacts of climate change. Marine and Freshwater Research 65(12), 1131–1144. DOI: 10.1071/MF14035. McJannet D, Yang A and Seo L (2023) Climate data characterisation for hydrological and agricultural scenario modelling across the Victoria, Roper and Southern Gulf catchments. A technical report from the CSIRO Victoria River and Southern Gulf Water Resource Assessments for the National Water Grid. CSIRO, Australia. McVicar T, Vleeshouwer, J., Van Niel, T., Guerschman, J. (2022) Actual Evapotranspiration for Australia using CMRSET algorithm. Version 1.0.0. Terrestrial Ecosystem Research Network (TERN). https://doi.org/10.25901/gg27-ck96. Merrin L, Stratford D, Kenyon R, Pritchard J, Linke S, Ponce Reyes R, Buckworth R, Castellazzi P, Costin B, Deng R, Gannon R, Gao S, Gilbey S, Lachish S, McGinness H and Waltham N (2024) Ecological assets of the Southern Gulf catchments to inform water resource assessments. A technical report from the CSIRO Southern Gulf Water Resource Assessment for the National Water Grid. CSIRO, Australia. Moya CE, Raiber M, Taulis M and Cox ME (2015) Hydrochemical evolution and groundwater flow processes in the Galilee and Eromanga basins, Great Artesian Basin, Australia: A multivariate statistical approach. Science of The Total Environment 508, 411–426. DOI: 10.1016/j.scitotenv.2014.11.099. Mueller N, Lewis A, Roberts D, Ring S, Melrose R, Sixsmith J, Lymburner L, McIntyre A, Tan P, Curnow S and Ip A (2016) Water observations from space: Mapping surface water from 25years of Landsat imagery across Australia. Remote Sensing of Environment 174, 341–352. DOI: https://doi.org/10.1016/j.rse.2015.11.003. Murphy R, Neal B, Morden R, Nathan RN and Evans R (2008) BaseJumper – tool for analysing timetrends in baseflow. In: Water Down Under, Adelaide, Engineers Australia. Norton CJ and Rollet N (2023) Regional stratigraphic correlation transects across the Great Artesian, Lake Eyre basins and Upper Darling Floodplain region (part 2): Infilling data and knowledge gaps. Record 2023/28. Geoscience Australia, Canberra. https://dx.doi.org/10.26186/147243. NT Government (2023) Territory Water Plan. A plan to deliver water security for all Territorians, now and into the future. Viewed 6 September 2024, https://watersecurity.nt.gov.au/__data/assets/pdf_file/0003/1247520/territory-water- plan.pdf. NVIS Technical Working Group (2017) Australian Vegetation Attribute Manual: National Vegetation Information System. Version 7.0. Department of Agriculture Water and the Environment, Canberra. Orr ML, Bernardel G, Owens R, Hall LS, Skeers N, Reese B and Woods M (2020) Geology of the Beetaloo GBA region. Technical appendix for the Geological and Bioregional Assessment: Stage 2. Department of the Environment and Energy, Bureau of Meteorology, CSIRO and Geoscience Australia, Australia. Available online: https://www.bioregionalassessments.gov.au/sites/default/files/gba-bee-stage2- appendix_geology_final.pdf. Owen N, Fan R and Delle Piane C (2023) XRD Report - Minerology of NDI Carrara 1 Samples. Australia G, Canberra. Pebesma EJ (2004) Multivariable geostatistics in S: the gstat package. Computers & Geosciences 30(7), 683–691. DOI: http://doi.org/10.1016/j.cageo.2004.03.012. Petersen-Øverleir A, Soot A and Reitan T (2009) Bayesian Rating Curve Inference as a Streamflow Data Quality Assessment Tool. Water Resources Management 23(9), 1835–1842. DOI: 10.1007/s11269-008-9354-5. Petheram C, Walker G, Grayson R, Thierfelder T and Zhang L (2002) Towards a framework for predicting impacts of land-use on recharge: 1. A review of recharge studies in Australia. Australian Journal of Soil Research 40(3), 397–417. Pettit NE, Bayliss P, Davies PM, Hamilton SK, Warfe DM, Bunn SE and Douglas MM (2011) Seasonal contrasts in carbon resources and ecological processes on a tropical floodplain. Freshwater Biology 56(6), 1047–1064. DOI: 10.1111/j.1365-2427.2010.02544.x. Pilgrim DH (1987) Australian Rainfall and Runoff: A Guide to Flood Estimation. The Institution of Engineers, Australia, Barton, ACT. Priestley SC, Shand P, Love AJ, Crossey LJ, Karlstrom KE, Keppel MN, Wohling DL and Rousseau- Gueutin P (2020) Hydrochemical variations of groundwater and spring discharge of the western Great Artesian Basin, Australia: implications for regional groundwater flow. Hydrogeology Journal 28(1), 263–278. DOI: 10.1007/s10040-019-02071-3. Queensland Government (2013) Salinity management handbook, 2nd editon. Department of Environment and Resources Management. Queensland Government (2021) Land use mapping – 1999 to 2019 – Queensland. Land use dataset downloaded 2 May 2023, https://www.qld.gov.au/environment/land/management/mapping/statewide- monitoring/qlump/qlump-datasets. Queensland Government (2022) Groundwater Database: Metadata, Data Dictionary & Standards Summary. State of Queensland. Exported on 10/01/2022. Current URL (HTML format): https://www.data.qld.gov.au/dataset/groundwater-database-queensland. Queensland Government (2023) Queensland Water Strategy. Water. Our life resource. Viewed 11 September 2024, https://www.rdmw.qld.gov.au/qld-water-strategy/strategic-direction. R Core Team (2016) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/. Radke BM, Ferguson J, Cresswell RG, Ransley TR and Habermehl MA (2000) Hydrochemistry and implied hydrodynamics of the Cadna-owie - Hooray Aquifer, Great Artesian Basin, Australia. Bureau of Rural Sciences, Canberra. Radke BM, Kellett JR, Ransley TR and Bell J (2012) Lexicon of the lithostratigraphic and hydrogeological units of the Great Artesian Basin and its Cenozoic cover. A technical report to the Australian Government from the CSIRO Great Artesian Basin Water Resource Assessment, Australian Government. CSIRO Water for a Healthy Country Flagship, Australia. Randal MA (1966) 1:125,00 Geological series - explanatory notes, Rankin, N.T. Sheet SE/53-16. Canberra. Raiber M, Lewis S, Cendón DI, Cui T, Cox ME, Gilfedder M and Rassam DW (2019) Significance of the connection between bedrock, alluvium and streams: A spatial and temporal hydrogeological and hydrogeochemical assessment from Queensland, Australia. Journal of Hydrology 569, 666–684. DOI: https://doi.org/10.1016/j.jhydrol.2018.12.020. Raiber M, Suckow A, Deslandes A, Gerber C, Martinez J, Crane P, Ransley T, Evans T, Wallace L and Wu G (2022) Assessing recharge processes and flow dynamics using environmental tracers in the Great Artesian Basin. CSIRO, Australia. Raiber M, Feitz A, Flook S, Cendón D, Suckow A, Schoening G, Hofmann H, Martinez J, Maas R and Kelly Bryce (2024) Strontium isotopes in the atmosphere, geosphere and hydrosphere: developing a systematic “fingerprinting” framework of rocks and water in sedimentary basins in eastern Australia. Science of Total Environment, Volume (951):1-20. https://doi.org/10.1016/j.scitotenv.2024.175522. Ransley TR and Smerdon BD (2012) Hydrostratigraphy, hydrogeology and system conceptualisation of the Great Artesian Basin. CSIRO Water for a Healthy Country Flagship, Australia. Ransley TR, Radke BM, Feitz AJ, Kellett JR, Owens R, Bell J, Stewart G and Carey H (2015) Hydrogeological Atlas of the Great Artesian Basin. Geoscience Australia, Canberra. Raymond OL, Liu SF, Kilgour P, Retter AJ, Stewart AJ and Stewart G (2007) Surface geology of Australia 1:1,000,000 scale, New South Wales - 2nd edition [Digital Dataset]. Geoscience Australia. http://www.ga.gov.au, Canberra. Raymond OL, Liu S, Gallagher R, Highet LM and Zhang W (2012) Surface Geology of Australia, 1:1 000 000 scale, 2012 edition [Digital Dataset]. Geoscience Australia, Commonwealth of Australia, Canberra., https://ecat.ga.gov.au/geonetwork/srv/eng/catalog.search. Read R (2003) Avon Downs and Ranken sheets groundwater occurrence. Northern Territory Department of Infrastructure, Planning and Environment, Natural Resources Division, Report 40/2003A. Scanlon BR, Keese KE, Flint AL, Flint LE, Gaye CB, Edmunds WM and Simmers I (2006) Global synthesis of groundwater recharge in semiarid and arid regions. Hydrological Processes 20(15), 3335–3370. Schofield KA, Alexander LC, Ridley CE, Vanderhoof MK, Fritz KM, Autrey BC, DeMeester JE, Kepner WG, Lane CR, Leibowitz SG and Pollard AI (2018) Biota connect aquatic habitats throughout freshwater ecosystem mosaics. Journal of the American Water Resources Association 54(2), 372–399. Schroder IF, de Caritat P, Wallace L., Trihey JM, Boreham C, Tobin J, English PE, Sohn J, Palatty P, Coghlan RA, Flower C, Maher P, Main PT, Thorose M, Long I, Byass J and Jinadasa N (2020) Northern Australia Hydrogeochemical Survey: Final Data Release and Hydrogeochemical Atlas for EFTF. Geoscience Australia, Canberra. http://dx.doi.org/10.11636/Record.2020.015. Smerdon BD, Ransley TR, Radke BM and Kellett JR (2012) Water resource assessment for the Great Artesian Basin. A report to the Australian Government from the CSIRO Great Artesian Basin Water Resource Assessment. CSIRO Water for a Healthy Country Flagship, Australia. Available online: https://doi.org/10.4225/08/584c45a39e1b5. Smith KG (1972) Stratigraphy of the Georgina Basin. Department of National Development, Bureau oF Mineral Resources, Geology and GEOPHYSICS. Bulletin No. 111. Available online: https://d28rz98at9flks.cloudfront.net/122/Bull_111.pdf. Smith JW and Roberts HG (1972) Mount Drumond, Northern Territory 1:1250,000 Geological series - explanatory notes. Record SE5312. Canberra. Suckow A (2014) The age of groundwater – definitions, models and why we do not need this term. Applied Geochemistry 50, 222–230. Tarantola A (2005) Inverse problem theory and methods for model parameter estimation. Society for Industrial and Applied Mathematics, Philadelphia, PA, USA. Taylor AR, Doble RC, Crosbie RS, Barry KE, Harrington GA, Davies PJ and Thomas M (2018a) Hydrogeological assessment of the Bulimba Formation – Mitchell catchment, Queensland. A technical report to the Australian Government from the CSIRO Northern Australia Water Resource Assessment, part of the National Water Infrastructure Development Fund: Water Resource Assessments. CSIRO, Australia. DOI: https://doi.org/10.25919/5b86ed86884f1. Taylor A, Doble R, Crosbie R, Barry K, Harrington G, Davies P and Thomas M (2018b) Hydrogeological assessment of the Bulimba Formation - Mitchell catchment, Queensland. A technical report to the Australian Government from the CSIRO Northern Australia Water Resource Assessment, part of the National Water Infrastructure Development Fund: Water Resource Assessments. CSIRO. Available online: https://doi.org/10.25919/5b86ed86884f1. Taylor A, Harrington G, Clohessy S, Dawes W, Crosbie R, Doble R, Wohling D, Batlle-Aguilar J, Davies P, Thomas M and Suckow A (2018c) Hydrogeological assessment of the Grant Group and Poole Sandstone - Fitzroy catchment, Western Australia. A technical report to the Australian Government from the CSIRO Northern Australia Water Resource Assessment, part of the National Water Infrastructure Development Fund: Water Resource Assessment. CSIRO. Available online: https://doi.org/10.25919/5b86ed7e26b48. Taylor AR, Barron OV, Harrington GA, Chilcott C, Davies PJ, Wohling DL and Gilfedder M (2021) Rapid appraisal of new groundwater-based irrigation infrastructure opportunities across Australia. A technical report from CSIRO to the National Water Grid Authority. CSIRO Technical Report EP2021-0769: CSIRO, Australia. Taylor AR, Turnadge C, Crosbie RS, Lamontagne S, Deslandes A, Davies PJ, Barry K, Suckow A, Knapton A, Marshall S, Hodgson G, Tickell S, Duvert C, Hutley L and Dooley K (2024) iHydrogeological assessment of the Tindall Limestone and equivalents, and the Dook Creek Formation and Nathan Group in the Roper catchment, Northern Territory. Thomas M, Philip S, Zund P, Stockmann U, Hill J, Gregory L, Watson I and Thomas E (2024) Soils and land suitability for the Southern Gulf catchments. A technical report from the CSIRO Southern Gulf Water Resource Assessment for the National Water Grid. CSIRO, Australia. Tickell SJ (2003) Water Resource Mapping Barkly Tablelands. Department of Infrastructure Planning and Environment Natural Systems Division. Report 23/2003, Darwin, September 2003. Available online: https://territorystories.nt.gov.au/10070/674454/0/0. Tickell SJ and Bruwer Q (2017) Georgina Basin Groundwater Assessment: Daly Waters to Tennant Creek, Technical Report 17/2017. Department of Environment and Natural Resources, Palmerston. Turnadge C, Crosbie RS, Tickell SJ, Zaar U, Smith SD, Dawes WR, Davies PJ, Harrington GA and Taylor AR (2018) Hydrogeological characterisation of the Mary–Wildman rivers area, Northern Territory. CSIRO, Australia. Available online: https://doi.org/10.25919/5b86edacde2ae. Twidale CR (1956) A physiographic reconnaissance of some Volcanic Provinces in North Queensland, Australia. Bulletin Volcanologique 18, 3-23. Vizy J and Rollet N (2023a) Eastern Australia 3D geological and hydrogeological surfaces update, Version 2.0: Report and data package of updated Great Artesian Basin and Lake Eyre Basin surfaces extending to additional onshore-offshore eastern Australian basins. Record 2023/44. Geoscience Australia, Canberra. Vizy J and Rollet N (2023b) Great Artesian Basin Vshale modelling, Version 1.0‒Data package of Vshale modelling based on the Eastern Australia 3D geological and hydrogeological surfaces update, Version 2.0. Geoscience Australia, Canberra. Vizy J and Rollet N (2024) Preliminary 3D geological model of the Georgina-southern McArthur- South Nicholson basins - Data package of 3D modelling geological surfaces and isochores, Version 1.0. Geoscience Australia, Canberra. https://dx.doi.org/10.26186/149290. Von Neumann J (1951) 13. various techniques used in connection with random digits. Appl. Math Ser 12(36–38), 5. Wallace L and Ransley T (eds) (2022) Assessing the Status of Groundwater in the Great Artesian Basin: Summary Report. Record 2022/46. Geoscience Australia, Canberra. Waltham N, Burrows D, Butler B, Wallace J, Thomas C, James C and Brodie J (2013) Waterhole ecology in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy, 460. Wilkins A, Crosbie R, Louth-Robins T, Davies P, Raiber M, Dawes W and Gao L (2022a) Australian gridded chloride deposition-rate dataset. Data in Brief 42, 108189. DOI: https://doi.org/10.1016/j.dib.2022.108189. Wilkins A, Crosbie RS, Raiber M, Dawes W, Louth-Robins T and Gao L (2022b) Quantifying recharge to groundwater systems in the NSW coalfields, ACARP Report C33037. CSIRO Report EP2022- 1381. Withnall IW and Cranfield LC (2013) Queensland geological framework. Queensland Minerals, 2013. Available online: https://www.resources.qld.gov.au/__data/assets/pdf_file/0007/197647/geology-of- queensland.pdf. Wood WW and Sanford WE (1995) Chemical and Isotopic Methods for Quantifying Ground-Water Recharge in a Regional, Semiarid Environment. Ground Water 33(3), 458–468. Wood WW (1999) Use and misuse of the chloride-mass balance method in estimating ground water recharge. Ground Water 37(1), 2–3. Xiang H, Zhang Y and Richardson JS (2016) Importance of Riparian Zone: Effects of Resource Availability at Land-water Interface. Riparian Ecol. Conserv, 2016(3), 1–17. Yin Foo D and Matthews I (2001) Hydrogeology of the Sturt Plateau. 1:250000 Scale Map Explanatory Notes. Department of Infrastructure, Planning and the Environment, Darwin., https://territorystories.nt.gov.au/10070/673614. As Australia’snational scienceagency and innovation catalyst, CSIRO is solving the greatestchallenges through innovativescience and technology. CSIRO. Unlocking a better futurefor everyone. Contact us 1300 363 400+61 3 9545 2176csiroenquiries@csiro.aucsiro.au For further informationEnvironment Dr Chris Chilcott+61 8 8944 8422chris.chilcott@csiro.au EnvironmentDr Cuan Petheram+61 3 6237 5669cuan.petheram@csiro.au Agriculture andFoodDr Ian Watson+61 7 4753 8606 Ian.watson@csiro.au