WHILE yield forecasting technology has been around for decades, new developments in satellite imagery has made it more accessible to growers.
One University of New England (UNE) research fellow believes the technology has the power to deliver substantial benefits to growers, industry and government.
Dr Andrew Robson first began working with satellite images 13 years ago, and in that time, he’s applied them to crops including peanuts, cotton, avocado, sugar cane and pastures.
With his UNE team, Dr Robson creates maps from satellite imagery.
Underpinning the maps are algorithms which convert imagery values to a range of agronomic
parameters such as tree and crop health and yield.
For growers, these maps clearly indicate regions of a crop that perform better within a growing season and across growing years.
“With this information coupled with co-ordinated field sampling, growers can develop a stronger understanding of crop variability, its drivers and its impact on yield and quality, thus supporting improved management and harvest strategies to optimise production,” Dr Robson said.
At the regional level, satellite imagery can accurately indicate location, total crop area planted as well as total production which can assist processors in their forward-marketing
operations.
It can also help governments to better implement biosecurity responses.
Satellite imagery also serves as an important technology for value adding the evolution of newer technologies such as robotics.
For the avocado industry, Dr Robson said the maps gave robotic engineers and growers the
perspective they needed to revolutionise their systems.
“It shows them how data at the individual leaf level can be extrapolated to the scale of whole farms and orchards,” he said.
“My work shows robotics students and practitioners they have to concentrate on developing tools that have a practical end use.”
Satellite imagery could be used to co-ordinate the work of autonomous machines, he said.
Where a yield map has identified poor tree or crop health, machines could be sent in to investigate the underlying cause.
At present, humans do that job.
“Robots could be used to calibrate the imagery – to ground truth the information the maps have identified.”
The avocado industry has increasingly used satellite imagery for yield forecasting, and Dr
Robson’s ongoing research with Simpson Farms, an avocado growing company based in
Childers, Queensland, is a good example.
The UNE research team with Simpson Farms’ senior agronomist Chad Simpson and researchers from The Department of Agriculture, Forestry and Fisheries, Qld, have used high resolution satellite imagery to classify trees according to their production capabilities
allowing the company to better manage trees and prioritise areas for harvest.