Aerial image of a forest fire

© Nebojsa |


Wildfires affect an estimated four percent of the global vegetated area each year. Fire-prone ecosystems are found across the world and climate change is expected to increase the risk of wildfires due to elevated temperatures and changes in the water cycle. Human society and economy are greatly impacted by fire due to the loss of lives and the destruction of livelihoods and critical infrastructure.

Earth observation provides an important tool for characterizing the impact of wildfires on society and ecosystems by providing spatial information on burned area and burn severity, on the one hand, and on environmental factors influencing fire hazard, such as fuel parameters (moisture, structure), land use/land cover, topography etc., on the other hand.

The CLIMERS research group is interested in quantifying the controls of climate, vegetation and human activity on wildfire occurrence and burned area. We aim at estimating fuel moisture using satellite datasets on soil moisture (e.g. CCI) and vegetation (e.g. VOD) and apply data-driven modelling methods to test for relations between these and other predictors and fire activity both at the global and regional scales.

Ongoing projects

Fire is an important feature of the ecology and diversity of terrestrial ecosystems and has accompanied humans on evolutionary time-scale. This project aims to improve the understanding of how humans influence fire occurrence and how fire occurrence and effects on ecosystems and society will change over the 21st century.

Read more on the FURNACES page

Past projects

Biomass burning is a globally significant source of aerosols, greenhouse gases and other trace gas species, influencing both regional and global climate. In this project, we explored multiple Earth observation datasets to infer relationships between soil moisture, fuel load production, and fire extent and emissions by using observations and machine learning techniques, and evaluated and improved state-of-the-art coupled dynamic global vegetation and fire models.

Read more on the CCi4SOFIE page

In recent years, the Alps and other mountain regions in Europe have been increasingly affected by forest fires. The aim of the project was to use novel remote sensing methods and state-of-the-art machine learning methods to develop daily analyses and forecasts of the ignitability and propagation risk of forest fires according to the requirements of meteorologists, fire brigades, forest specialists and infrastructure providers.

Read more on the CONFIRM page