Natural England has used an AI model to develop a highly detailed map of peatland across the country.
It has completed the AI4Peat project, working with partners include the National Trust, Moors for the Future and the Yorkshire Peat Partnership.
The project was prompted by the lack of a detailed map showing important features such as man-made drainage, natural channels and existing dams, which affect peatlands’ ability to absorb water, which in turn undermines its capacity to absorb carbon from the atmosphere.
AI4Peat, which began in 2021, has combined aerial photography with two AI models: semantic segmentation to detect grips, gullies and hags, distinguishing them from the surrounding landscape; and object detection to identify dams to track where restoration has already occurred.
They were trained using existing digitised data and LiDAR terrain data and deployed over all areas of England’s upland peatland.
Tech resources
The project also utilised the Department for Environment, Food and Rural Affairs’ (Defra) Analytics and Science Hub, the Microsoft Azure platform and Databricks technology. These have made it possible to detect the shape size and boundaries of drainage features in images of the peatland.
Natural England said it has succeeded in processing the images into the England Peat Map which will be made available during the spring. It will also show the extent, depth and vegetation coverage of peat and peaty soils.
“The success of this approach demonstrates how AI can be used to improve the efficiency of environmental monitoring,” said senior project manager Anne Williams and Defra senior data scientist Martha Tabor.
“These types of techniques can be replicated, for example to detect or count animal species of interest, or to track changes in land use or monitor coastal erosion, provide vital insights into how our natural environment is changing and whether policy interventions are working as expected.”
