AI a tool for nitrogen decisions
Artificial intelligence could soon help grain growers know just how much nitrogen fertiliser to apply, according to a leading CSIRO researcher.
Roger Lawes is one of dozens of speakers set to take to the stage at the Grains Research and Development grains research updates in Perth on February 25 and 26.
The farming system scientist will also speak at the GRDC’s regional grains research update at Kendenup Lodge on February 28.
Dr Lawes said research showed on-farm strip trials, combined with crop modelling and satellite imagery analysis, could help growers understand whether a crop in a particular paddock would respond to an application of nitrogen, regardless of crop type, soil or season.
He said conventional crop nitrogen fertiliser decision-making tools were complex and required information about the soil’s nitrogen content and crop requirements.
“Growers need to measure the starting amount of nitrogen present in the soil, accounting for mineralisation,” he said.
“They then have to estimate the requirements for nitrogen by the crop.
“These factors vary with the season and soil type, complicating decision making.”
Dr Lawes said CSIRO had used GRDC investment to develop a framework to help modernise and simplify growers’ nitrogen decisions.
The framework used machine learning, remote sensing, on-farm trialling and crop modelling to work out whether a crop would respond to a nitrogen application.
“We found that modern analytical approaches, combined with on-farm experimentation and the sensing of multiple crops, may enhance nitrogen fertiliser recommendations for wheat crops,” Dr Lawes said.
He said previous research by CSIRO had found applying nitrogen according to the long-term mean yield of a paddock could provide an “economically-sensible approach to nitrogen fertiliser management”.
“Since that study, artificial intelligence and machine learning techniques have evolved,” he said.
“It is now possible to process large volumes of information about crops, using satellite imagery or information generated from crop models like the Agricultural Production Systems sIMulator.”
Dr Lawes said the study aimed to identify the variables that needed to be monitored to help growers make a profitable decision.
“The machine learning technique ‘Random Forest’ was used to determine which variables were the most important and useful for predicting the optimal nitrogen fertiliser rate for a paddock,” he said.
“As expected, this analysis found the most important variable was the long-term historical mean site yield, as estimated from APSIM, the crop simulation model used in this study.
“Extractable soil water to a depth of 150cm and leaf nitrogen content determined by the ‘nitrogen minus strip’ were the two next most important variables.”
CSIRO launched a new artificial-intelligence-powered platform last month, bringing together a trove of land-use data for agribusinesses to predict performance.
The Rural Intelligence Platform combined a variety of technology developed by CSIRO, including a digital soil map and satellite imagery analysis.
The platform uses machine learning to create a relatively accurate estimate of a residential property’s value with what the CSIRO says is 90 per cent accuracy.
To find out more, visit giwa.org.au/2019researchupdates or grdc.com.au/updatedates.
Get the latest news from thewest.com.au in your inbox.
Sign up for our emails