Scientists have a new way to keep watch over Antarctica's smallest inhabitants. A research team led by the University of Wollongong combined drones, satellite imagery and artificial intelligence to map the continent's scattered mosses, lichens and other tiny organisms, according to a study published in the ISPRS Journal of Photogrammetry and Remote Sensing.
The approach targets a problem that has frustrated researchers for decades. Antarctica's plant life does not grow in vast, easy-to-spot forests. It clings to isolated rocky outcrops, survives in thin patches, and often hides under snow. That makes it expensive and dangerous for scientists to find and monitor on foot.
The new method puts drones to work flying over those hard-to-reach areas. Their high-resolution images get fed into AI models trained to recognize moss beds, lichen communities and other small ecosystems. Satellites then fill in the bigger picture, tracking changes across far larger stretches of the continent than a single drone flight could cover.
"This gives us a repeatable, scalable way to observe life in one of the most remote environments on Earth," the researchers said in describing the work, which pairs remote sensing with machine learning.
Why does it matter? Antarctica's small lifeforms sit at the base of a food web that supports insects, mites and birds. They also respond quickly to temperature shifts, making them sensitive early indicators of climate change. If moss beds shrink or move, that signals something larger happening in the ecosystem.
Until now, tracking those changes meant sending teams into the field season after season, often to the same remote sites. The UOW-led approach could cut that workload. Drones can cover ground quickly, and the AI handles the tedious job of sorting through thousands of images to find patches of green.
The study's authors say the technique is not a replacement for on-the-ground science. It is a tool to decide where researchers should look next and to catch changes between expeditions. The team tested the method on real Antarctic terrain, and the results suggest it can reliably detect the small, scattered plant communities that traditional surveys miss.
Antarctica is warming faster than most places on Earth, and its ice-free areas, where almost all of its plant life grows, are expanding. That makes accurate, frequent monitoring more urgent. The new system could give scientists a running record of how life responds as the continent changes.
The research was published in the ISPRS Journal of Photogrammetry and Remote Sensing. The team says future work will focus on expanding drone coverage and improving the AI's ability to tell one species from another from the air.
