FUELITY

Aim
This project aims to provide a comprehensive, long-term analysis of fuel characteristics to understand the role of vegetation in controlling the changing trends of fire activity on the planet. To achieve this goal, Fuelity will leverage the extensive information available from ESA-funded missions that have improved biomass estimations. It will also complement several other projects, primarily EO-based, and build upon ongoing initiatives like SMOS and planned ESA missions such as Biomass and Flex, which are designed to enhance our understanding of vegetation-related factors.

Why It’s Important
Monitoring fuel status is of paramount importance because the carbon released during fires offsets the crucial carbon sequestration performed by vegetation. Furthermore, fuel availability drives fire activity, and fuel status determines fire emissions. Despite the significance of these Earth system components, methods to monitor their evolution are still lacking. This is partly because the information is often intertwined, with most Earth Observation (EO) data, such as LAI and VOD, contributing to both fuel load and fuel status. While several methods have been developed in recent years using in situ and remote observations to monitor certain components of vegetation status, an EO-based comprehensive dataset that includes both dead and live components for mass and moisture is, to the best of our knowledge, still missing.

How Fuelity is New
In the past year, ECMWF has achieved a significant breakthrough in real-time fuel status predictions by implementing a new vegetation characteristics model (referred to hereafter as the “fuel model”). This scheme provides frequent daily updates on vegetation characteristics at a high spatial resolution (9 km). These characteristics encompass more than just load and moisture; they also include detailed attributions for foliage and woody components of both live and dead vegetation. The predictive accuracy of this data could be significantly enhanced by incorporating EO observations into the system. The most innovative aspect of this project is, therefore, to advance the assimilation component, aiming to deliver an EO-informed, comprehensive long-term analysis of fuel characteristics.

What Will It Bring?
At the project’s conclusion, we envision the release of a significant database to the scientific community, which will serve as a fundamental asset for understanding how changing patterns of fuel availability are impacting fire regimes worldwide. The dataset will be released in an AI-ready format to facilitate its utilization by machine learning scientists. Several downstream applications are possible, including the development of new fire danger indices and the enhancement of global fire emissions estimates.

What Else?
Moreover, the established infrastructure for initializing model predictions with available observations is considered an initial step toward implementing real-time analysis of fuel load and fuel moisture. Monitoring these critical climate variables represents a significant advancement in our understanding of changing patterns in fire activity worldwide and in refining global fire emission estimates.