Monitoring crop growth from the skies

Aerial hyperspectral imaging and modelling techniques in development at the University of Melbourne have allowed researchers to rapidly measure the individual health of each tree in a 1000-hectare orchard.

This new approach to monitoring crop health is expected to improve crop productivity, leading to higher yields, and the efficient use of crop inputs like fertilisers and water, resulting in benefits for the environment and improving profitability for growers.

View of crops taken from air with wing of light aircraft visible
Aerial hyperspectral imaging and modelling techniques are being developed to help growers better evaluate stress on their crops

Professor Pablo Zarco-Tejada and his team in the HyperSens Remote Sensing Laboratory combine the expertise of the Faculty of Veterinary and Agricultural Sciences and the Melbourne School of Engineering.

In February 2020, the team used a light manned aircraft to gather images and create effective stress maps for a large almond orchard in the Mallee region. A larger-scale project is now in development with the Almond Board of Australia to expand this research.

While drones have previously been used to establish hyperspectral images and modelling techniques, their operational capacity limits them to smaller scale projects.

Small-scale drone examples include the development of a disease database for tomato crops and water stress signatures in stone fruit crops.

Professor Zarco-Tejada says the enhanced airborne facility established at the University has made it possible to scale up this research over larger areas to include broadacre crops.

His plane is fitted with thermal and hyperspectral camera equipment to capture images at a fine-scale resolution to identify stress factors like water and nutrient deficiencies in individual plants or at scale.

The facility is operated by Dr Tomas Poblete, Research Fellow (Airborne Remote Sensing), who is in charge of camera installation and operation of the instruments on board the aircraft.

View from ground of freshly shaken almonds with blurred almond trees in background
The trial mapped a 1000-hectare Victorian almond orchard

The trial mapped a 1000-hectare Victorian almond orchard, providing the grower with detailed stress maps within 24 hours of the imaging flight.

MSE PhD student Anne Wang and Postdoctoral Research Fellow Dr Lola Suarez led the work on the ground, gathering data about leaf pigment, soil and vegetation reflectance based on temperatures from the field.

Combining the airborne imagery with the ground measurements has quantified stress levels for every tree in the orchard. The resulting maps show chlorophyll and nitrogen levels from the hyperspectral images, as well as water stress from thermal imagery.

Professor Zarco-Tejada says the maps allow the grower to see variability in the stress indicators across the whole orchard and to understand why yield losses are occurring in some areas.

“This new method of mapping allows the grower to consider variable applications of water and/or nutrients to better account for different soils across the orchard and the needs of trees, given varying ages, dimensions and vigour,” he says.

Show of rows of vines in vineyard with blue sky in background
Professor Zarco-Tejada hopes the team's efforts will expand to include citrus orchards and vineyards in the future

When matched with tree-by-tree yield data from the grower, the research team can quantify the effects of the mapped stressors on crop yields. This information allows the grower to make cost-benefit analysis of variable water and nutrient applications.

This project shows the potential of smart agricultural systems integrating larger, longer-range aircraft and computer modelling to provide actionable insight across agriculture, says Associate Professor Davi La Ferla, Enterprise Fellow (Smart Agriculture) in the Faculty of Veterinary and Agricultural Sciences and CEO of digital agriculture company Sensand Technologies.

“The challenges and opportunities primary producers and land managers face across Australia are becoming clearer in light of our changing and more dynamic environments,” he says.

“Agricultural productivity, operational profitability, biosecurity, environmental stewardship and biodiversity are often top of mind.

“Recent advancements in plant science, animal science, soil science, computer science and social science have enormous potential to transform Australian agriculture, underpin emerging economies and secure our future.

“I am increasingly confident that the integration of Professor Zarco-Tejada’s leading-edge remote sensing with other promising agricultural technologies will be an important driver of this change.”

The HyperSens team’s research is part of a network of projects receiving funding from the Mallee Regional Innovation Centre (MRIC) in Mildura. MRIC is an applied research collaboration bringing together local business, the University of Melbourne, LaTrobe University and SuniTAFE.

Reflecting on the potential of the project, Professor Zarco-Tejada mused that he hopes to expand his team’s work to include citrus orchards, and possibly vineyards as well. We’ll drink to that!

For more information or to collaborate on the research, contact Professor Pablo Zarco-Tejada.

This story first appeared on the Melbourne School of Engineering’s Water, Environment and Agriculture industry news page.