boxed-arrow-leftArtboard 1Artboard 1Artboard 1Artboard 1
Open menu

Published: 19 Mar 2021

The use of satellite imagery to rapidly monitor and increase pasture efficiency may become an easy to access tech tool for farmers, thanks to recent research at the Tasmanian Institute of Agriculture (TIA) at the University of Tasmania.

As part of the two-year project, led by TIA’s internationally renowned agricultural systems modelling scientist Associate Professor Matthew Harrison and Dr Bethany Melville from the University’s School of Geography, Planning and Spatial Sciences, 10 m multispectral imagery has been reliably used to distinguish between pasture species and seasonal changes in botanical composition in pastures.

Associate Professor Harrison said the results of the research showed great potential for efficiency improvements in livestock grazing enterprises - particularly for larger properties.

“For effective use of pastures, livestock grazing systems require real time pasture monitoring and measurement, however manual and physical monitoring of large areas of pasture is often expensive, time consuming and infeasible,” he said.

“The use of freely-available daily volumes of ‘big data’ from earth observation sources has the potential to aid real time management of pastures remotely, saving livestock managers both time and money.”

Based on a 1000 ha merino sheep property on the State’s East Coast, the research used daily high resolution multispectral (RGM and NIR) images from American private earth imaging company Planet Labs, to provide real-time monitoring to aid grazing management decisions.

And the results were very promising, showing satellite imagery could predict pasture classes with more than 75 per cent accuracy.

Associate Professor Harrison, whose expertise includes climate change impacts and adaptation of livestock and cropping enterprises, said the research had positive implications for maintaining and improving agricultural sustainability over the long-term.

“This shows exciting potential for using satellite imagery to support grazing management decisions to both benefit pasture production and improve environmental stewardship,” he said.

While other attempts to use satellite imagery to monitor pastures had been made in the past, Associate Professor Harrison said the results had not yielded the results necessary to assist farmers.

“Past satellite imagery has neither had the spatial resolution (pixel size) nor pass-rate (number of satellite flyovers) to enable effective decision making in grazing management at the farm or paddock levels,” he said.

“This work uses the latest high-resolution high frequency satellite constellations in concert with advanced artificial intelligence and machine learning techniques, to enable timely and reliable decision-making at the paddock level.

“This will provide farm managers with more timely data on remaining pasture biomass to allow subsequent decisions, such as movement of livestock or the need for fertiliser inputs, irrigation or supplementary feeding.”

Researchers are now working on linking the model to an agtech app and testing the viability of the approach with Tasmanian farmers.

It is hoped the model will be available more broadly in 2022.

The research project was supported by Smart Farming Partnerships provided in the second phase of the National Landcare Program, delivered by the Department of Agriculture, Water and the Environment (DAWE).

TIA is a joint venture between the University of Tasmania and the Tasmanian Government.