A combination of satellite imagery, environmental data and habitat surveys carried out by citizen scientists could help transform the way endangered habitats are mapped, according to new research from the University of Surrey and Surrey Wildlife Trust.
As biodiversity declines and many of the UK's most important habitats become smaller and more fragmented, conservation organizations need accurate, up-to-date maps to protect wildlife and restore nature. However, creating these maps by surveying every site on foot is time-consuming, expensive and often impractical across large areas.
Working in collaboration with Surrey Wildlife Trust, Buglife and the Painshill Park Trust as part of the Space4Nature project, scientists from Surrey researched how citizen science, satellite imagery and machine learning could be used to produce detailed maps of lowland heathland – one of the UK's rarest and most threatened habitats.
In the study, published in People and Nature, trained volunteers used a mobile app to carry out hundreds of surveys across two of Surrey’s heathland sites – Chobham Common and Puttenham Common. The research team combined those observations with satellite imagery, soil and landscape data to develop a machine learning model capable of predicting where lowland heathland is likely to occur across the county.
The resulting habitat map was produced at a resolution of just three meters, helping conservationists to identify small pockets of heathland that can often be missed by broader national mapping products. The approach could provide conservation organizations with a faster, more cost-effective way to monitor habitats, target restoration efforts and better protect biodiversity.
Dr. Ben Siggery, Research and Monitoring Manager at Surrey Wildlife Trust, said:
"Protecting nature starts with knowing where it is, but creating detailed habitat maps has traditionally required huge amounts of time and resources. Our research shows that, with the right training and support, citizen scientists can collect reliable ecological data that can be combined with satellite imagery and machine learning to help us map habitats more efficiently."
Alongside reducing the time and cost needed to map habitats, researchers say the approach strengthens public engagement in conservation by enabling volunteers to contribute directly to scientific research and nature recovery.
Dr. Ana Andries, Project Co-Investigator on the study and Senior Lecturer in GIS and Remote Sensing at the University of Surrey, said:
“Our work shows what's possible when academia, conservation organizations and local communities work together. Combining satellite technology, machine learning and the enthusiasm of citizen scientists, we've developed an approach that could help conservationists monitor vulnerable habitats more quickly, direct restoration efforts where they're needed most and support biodiversity recovery at a much larger scale."
The approach is already proving transferable. Alongside this study, the research team has successfully applied the same framework to map Surrey’s fragmented chalk grassland, demonstrating how the combination of citizen science, Earth observation and expert validation provides a repeatable approach for updating habitat evidence beyond the limits of traditional field surveys.
Researchers say the framework could also be adapted for use across the UK and overseas, helping organizations target habitat restoration and reconnect fragmented landscapes.