Machine learning helps to map invasive gamba grass from space

Researchers from The University of Western Australia, CSIRO and Charles Darwin University have developed a machine-learning approach that reliably detects invasive gamba grass from high-resolution satellite imagery. 

Gamba grass, originally from Africa, is a weed of national significance, and one of five introduced grass species that poses an extensive and significant threat to Australia’s biodiversity. The perennial grass can grow to four metres in height and forms dense tussocks, which can burn as large, hot fires late in the dry season.

“Mapping gamba grass using satellite imagery unlocks the potential to frequently map large areas so we can get a better picture of where gamba grass is across northern Australia, and how quickly it is spreading."

Associate Professor Dr Samantha Setterfield

Associate Professor Dr Samantha Setterfield from UWA’s School of Biological Sciences said that accurate maps of where gamba grass occurred were essential to control the spread of the weed. 

“Mapping gamba grass using satellite imagery unlocks the potential to frequently map large areas so we can get a better picture of where gamba grass is across northern Australia, and how quickly it is spreading,” Dr Setterfield said.

“Managers can then target areas that are the highest priority for control, such as biodiversity-rich areas or culturally important sites.”

Dr Shaun Levick from Australia’s national science agency CSIRO said the research team used field data to train a machine-learning model to detect gamba grass from high-resolution, multispectral satellite imagery.

“Under optimum conditions, our method can detect gamba grass presence with about 90 per cent accuracy,” Dr Levick said.

The researchers commissioned the WorldView-3 satellite to capture very high-resolution imagery across 16 spectral bands for an area of 205 square kilometres near Batchelor in the Northern Territory – an area of dense gamba grass infestation. 

The wide range of data allowed them to use factors unseen to the human eye, such as leaf moisture levels and chlorophyll content, to differentiate between gamba grass and native grass species. 

Dr Natalie Rossiter-Rachor, of Charles Darwin University, said the project drew on extensive on-ground research into the life cycle of gamba grass to help achieve such accurate detection rates.

“We knew that gamba grass tends to stay green longer into the dry season than native grasses, so we timed the capture of the satellite imagery for this period,” she said.

“Understanding the ecology of the problem was essential to informing the remote sensing and machine-learning solution to the problem.”

The project, funded by the Federal Government’s National Environmental Science Program under the Northern Australia Environmental Resources Hub, is part of a larger effort to detect and map gamba grass throughout the north. 

The longer-term goal of the team is to move to a system where they can use free, open-access imagery to map gamba grass. They plan to develop a technique that is accessible to anyone and can help improve land management in northern Australia.

Read the paper Leveraging High-Resolution Satellite Imagery and Gradient Boosting for Invasive Weed Mapping.