AI for Geo-spatial Data Integration

Location: Ireland
Institution: Maynooth University, Trinity College Dublin
Status: Active
Type: Independent
Theme: Evidence Discovery & Integration
Timeframe: 2024 - 2029

Share on social

For decades, satellites have been collecting images of the Earth from space, building up a vast archive of environmental data. But making sense of all that information is not straightforward. Cloud cover can block satellite cameras, and processing radar data requires specialist skills. This project is developing new ways to overcome these barriers, combining data from satellites, aircraft, drones and ground-based sensors into a clearer, more complete picture of our environment.

At the heart of the project is machine learning, a type of artificial intelligence (AI) that can be trained to recognise patterns in large and complex datasets. By applying these techniques to environmental data, the project aims to map land cover and land use change, measure the extent and health of ecosystems, and assess how vulnerable our coastlines are to change.

The project is working closely with key organisations including the Environmental Protection Agency and Teagasc, and is focusing on a range of test locations across Ireland, from productive grasslands in Wexford and Kerry, to mixed land use areas in Mayo and Kildare, and croplands in Oakpark.

The results will include regularly updated maps and datasets that can be used for environmental reporting, ecosystem accounting, and informing the decisions needed to protect and restore Ireland’s natural environment.

Project Goals

Build a central data library by gathering and organising satellite images, aerial photography, drone footage, and ground-based sensor data into a single accessible platform, complete with clear descriptions so the data can be easily found, viewed and used.
Create detailed land cover maps by collecting new information on how different types of land across Ireland, including croplands, grasslands, wetlands and coastal areas, change over time and across seasons.
Test and improve AI mapping tools by assessing how well machine learning can identify and classify different types of land cover and develop approaches that outperform existing methods.
Combine different types of satellite data and explore how optical imagery and radar data can be used together to get a complete and more reliable picture of the landscape, even in cloudy conditions.
Develop an ecosystem mapping system by designing new automated tools that can measure and monitor the extent and health of ecosystems across Ireland, using satellite and sensor data in combination.
Refine and improve land monitoring techniques by developing advanced data analysis methods tailored to specific land cover types and use these to better understand how climate change is affecting the vulnerability of Ireland’s coastlines.