Digiscape-the-Science-David-Deery

David Deery  

 

 

[Image of a split circle appears with photos in each half of the circle flashing through of various CSIRO activities and the circle then morphs into the CSIRO logo]

 

[Image changes to show a new slide showing a photo of a crop on the left, and a furrow irrigation channel on the right, and David Deery can be seen inset in the top right talking, and text appears: Quantifying crop water use from field to regional scales using multi-resolution thermal sensing]

David Deery: So when talking about crop water use this is a, yeah, this is a cotton field and we’re essentially talking about the water that’s, you know, the evapotranspiration from the crop here. And then you’ve got to talk a little bit about some thermal sensing and we’ve got thermal sensing at different scales and what we, this is an example of static thermal sensing high-frequency measuring the temperature continuously.

 

[Image changes to show a new slide showing a line graph showing Annual allocation price scenarios for the Murrumbidgee River, and photos appear on the right of a Smart meter on a pole, and an irrigation sluice, and text appears: Cost of irrigation water, Cotton, ~6ML/ha -> 50ha ->, $30k to ~ $50k]

 

And so with irrigated cropping, over here this is furrow irrigation, one of the things is that there’s a handle, I suppose, somewhere, about the amount of water that is put onto the crop like either the smart metering is, I’m no expert but there is Smart metering out there, and then so we know that component, but we don’t really have a handle on how much water is actually transpired by the crop. And so one of the other aspects of this of course is the actual price of the water, of irrigation water so we’re looking here at the time series of the price per megalitre of water and you can see it’s a fairly variable thing and there’s a couple of different scenarios here looking forward. Essentially, I guess one of the take homes from this is that this is likely to probably increase through time and also with global warming the evaporative demand is going to increase. One of the other key pieces from this is, if we look at a small case study here of a cotton crop at several megalitres per hectare per year over 40 hectares the spend there the equal cost is in the order of tens of thousands of dollars.

 

[Image changes to show a new slide and David can be seen inset in the top right talking and text appears on the slide: What if we could quantify crop water use multiple scales (field/farm/region).., What would be the benefit?, Impact pathway workshop in June ’20 identified two potential applications, 1. Paddock and farm scale – enable corporate farms/multinationals to benchmark their water use as a social license to operate, 2. Catchment scale water productivity – water managers, benchmarking within and across commodities, Multi-disciplinary -> L&W satellite, A&F crop/physiology/digital and domain, Facilitated by CSIRO Corporate Strategy Team]

 

So we asked ourselves, OK so what if we could, what if we could quantify crop water use, global scale, so at a field-scale, paddock-scale, at the farm scale and across an irrigation region so if we could do it, what would be the benefit? So we sat down, had a little workshop a while back and we came up with two potential applications. So one is at the paddock and farm scale to enable corporate farms, multinationals to benchmark their water use in the sort of guise of a social license to operate, and the other one is, another one is at a catchment scale productivity, so for water managers, benchmarking within a commodity and, you know, across regions.

 

[Image changes to show a new slide showing a photo of a Smart meter on the Digiscape Future Science Platform website, and a photo of a cotton crop and canopy temperature graph appears on the right, and text appears: Pilot activity within WaterWise]

 

So then, just stepping back a bit, so this activity sits within WaterWise and what we, the, the piece here is we have thermal sensing and the high frequency continuous then Cloud-based Smart like data analytics and then advising growers on when the best time is to irrigate to make use of that resource and maximise their productivity.

 

[Image changes to show a new slide showing a diagram of how a plant takes in water, and how it moves through the plant system by evapotranspiration beneath the text heading: Science principles -> application]

 

So then in the science sense, around why we would use thermal, well, we got here is an application of this where, you know, plants transpire water and in exchange for carbon and so there’s this little bit of cooling that happens here and so we’re able to use that as the signal if you like of how stressed the plant is or not and then around using the [03.42] to irrigate but there’s a bunch of variables going on here with the, all of the weather pieces in this puzzle.

 

[Image changes to show a new slide showing photos of a cotton crop, a helicopter, a drone, a couple of diagrams, and a Landsat image, and text appears: Scale-up and quantify mm of water, Thermal measures at multiple temporal and spatial scales, What’s the value? To who and how?, Airborne thermal (on-demand, 10cm pixel size), Landsat 8 (every 16 days, 100m pixel size)]

 

So what we’re thinking about is, like, if we’re going to scale-up to quantify the millimetres of water, remembering that the growers they’ll pay, they pay for their water in, you know, it’s a metric of volume, so millimetres, right, across an area. The other, so what pieces have we’ve got in CSIRO? So we’ve got our high-frequency sensing here. We’ve got other little bits and pieces where we use this, this reference, as a non-transpiring lead to try and estimate the canopy water use then scaling up to another level with the thermal sensing from, essentially, these things are hooks in the sky that capture the thermal signature across a large area at a single point in time in the order of tens of centimetres. And then we scale up to the satellite where we’re looking at a frequency of around 16 days with 100m or thereabouts pixel size. And so then I, so then we’re also working out what’s the value to who and to how? So we have this thermal-sensing model to go on spatial scales.

 

[Image changes to show a new slide showing photos of a satellite map, a Landsat satellite, a cotton crop, a Smart meter in a crop, and a helicopter, and David can be see inset talking in the top right and text appears: Developing proof of concept, Data layers]

 

We’ve gone around and we’re doing a small proof-of-concept experiment this summer just completed so this is on a large cotton field here on the, in the Murrumbidgee irrigation area of Coleambally so we have a cotton large, like production cotton system, 50 hectares and we’ve got this is eco-variance for, there’s a gold standard benchmark for quantifying the crop water use, and then we’ve got data layers of thermal data at different scales. So, we have the high frequency data here and then we’ve got the, the thermal data at the same, with the time to have at the same time as the satellite overpass. And so just a little bit on the data from satellite.

 

[Image changes to show a new slide showing an image of the Landsat 8 satellite moving over the Earth’s surface, and inset photos of Tom and Tim can be seen inset at the bottom right and David can be seen inset talking in the top right, and text appears: Data from satellite, Landsat 8 (every 16 days, 100m pixel size)]

 

This is work from Tom Van Neil and Tim McVicar in CSIRO, Land and Water.

 

[Image changes to show a new slide showing two Google Earth evapotranspiration Australian maps from October 2020 and February 2021, and text appears: ETa for Australia, Monthly, 500mm pixel, Too coarse]

 

So what we have is, they have a, actual evapotranspiration model for the whole of Australia that they, that’s published monthly. This is at a 500m pixel resolution which is a little bit, probably too coarse for some of the field-scale work that we’re interested in and you can see here, these are two panels upwards from the model from October 2020 and February 2021 and the, yeah, so the bluer hues are ETing around 4mm per day and you can see here that we can’t really make out our field but yeah the point is that there is a model right for the whole of Australia. So we want to try and adapt that if you like to a field scale.

 

[Image changes to show a new slide showing a Landsat Surface Temperature image, and David can be seen inset on the top right talking to the camera, and text appears: Landsat Surface Temperature (every 16 days), 100m pixel size]

 

And what we used, and that is using the Micro satellite product and then this one this is using the Landsat product and that has a, this is a panel surface temperature every 16 days and the pixel size here is around 100m which is a little bit more amenable to what we’re interested in. So these redder hues, the temperature is greater than 35^OC and these blue, this one here is around 20^oC. So each of these little rectangular square shapes is an irrigated field.

 

[Image changes to show a new slide showing the colour image of a paddock on the left, and then a thermal image on the right showing the degrees of the paddock and text appears: Our paddock of interest near Carathool, ~3 deg C range in temperature across paddock]

 

And so this is our paddock of interest, I’ll show this, this here are at Carathool at the 9^th of February just zooming in here this is the colour image of that and then this is the thermal image of that so yeah. That’s you can see that there’s a couple of degrees sort of spatial resolution across there.

 

[Image changes to show a new slide and David can be seen inset in the top right talking and text appears: Challenges, future possibilities, Satellite overpass -> 16 days, Cloud compromises the data, Interpolation between data layers?, Scaled-up implementation will likely require other CSIRO tech., e.g. ePaddocks, Crop ID, Plant available water for dry land cropping (digital soil map)?]

 

And OK, so we’ve just got the final slide here, just a little bit around the challenges, future possibilities if you like. So one of the challenges with the satellite, is there’s an overpass of around 16 days and there’s the cloud compromises the data. And so if that thing goes, if there’s cloud on the day that the satellite goes over then you miss that and then that then blows out to a 32-day thing and if there’s two in a row then you have a 48-day frequency. So we’ve got to work out a way, if we’re going to, if this is a worthwhile thing to get going to, how do you interpolate between the data layers?

 

And then there’s other data modalities if you like, or data types and technologies that will be required for any kind of scale up here with things like the ePaddocks to work out the co-ordination of the paddock and the Crop ID. The last one just is that, yeah, like if we’re going to use this as a sort of a water balance in dry land cropping we’re going to have to bring in other things like the digital soil map and so on and so forth.

 

[Image changes to show a new slide showing text: Team Members, Warren Jin, Tony Nadelko, Akram Hameed, Bangyou Zheng, Tracey May, Simon Allen, Gordon McLachlan, David Smith, Rose Brodrick, Hiz Jamali, Chris Nunn, Quanxi Shao, Humaira Sultana, Tim McVicar, Tom Van Niel, Roy Zandona, Mike Caccetta]

 

Yeah, so thank you and just to make mention of the whole WaterWise team there. Thanks folks.

 

[Image changes to show a white screen and the CSIRO logo and text appears: CSIRO, Australia’s National Science Agency]