Dashboards to assist rice growers with decision making

The NSW Department of Industries project 'Agronomy and remote sensing to maximise rice water productivity' is using results from numerous variety and nitrogen experiments, along with commercial field data, to create and update growing guides, and panicle initiation (PI) and flowering date models.

The data from the NSW DPI project is being fed to the University of New England who are creating the dashboards for growers.

Associate Professor Dr James Brinkhoff, Applied Agricultural Remote Sensing Centre, UNE, is combining the results with weather and satellite data to create the real-time rice decision support tool.

"They are updated every day with latest weather and satellite data," he said.

"The DPI over the last however many years has collected thousand and thousands of data points on PI, flowering dates, yield, nitrogen.

"We've been able to bring that data set together with machine learning, satellite images, weather to try and predict in real time what is happening in every single field in the industry - this year there's just under 2000 fields."

The dashboards include phenology prediction, growth curves and grain moisture dry-down predictions.

Dr Brinkhoff said the phenology prediction included permanent water PI and flowering.

"It's really important to predict permanent water because that's what drives when you get to PI, when you get to flowering," he said.

Dr Brinkhoff said this season the accuracy was about 4.5 days.

"That's actually pretty good because the satellite goes over every five days and you would've seen it was really cloudy in November and December.

Dr Brinkhoff said the PI prediction accuracy was at 2.3 days, which given the warm weather at key timings was also good, while flowering accuracy was about five days.

"As we try and predict the later phenological stages using weather data, more things can creep in like nitrogen levels," he said.

"This year it looks like there's been a very high levels of nitrogen uptake on average and that tends to push flowering back a little bit that may have affected that as well."

The growth curves in the dashboard track what is happening in each paddock over time, which is benchmarked with the industry average.

"The curve indicates things like nitrogen status, and vigour but also we've seen cases where you can monitor water stress through it as well," he said.

"We've had spatial information for quite a while just looking at NDVI maps and variability.

"This is adding timing information and seeing how things are progressing through the season, so hopefully this ends up being a really useful decision support tool."

Dr Brinkhoff said they were taking moisture samples across several areas for grain moisture dry-down prediction.

"We're not just predicting harvest date," he said.

"We also predicting where's the grain moisture going to be next week or in a months time, so that might guide drainage decisions as well as the harvest decisions."

The dashboards have been rolled out for all growers, who can access it through Rice Extension.