Work

SEVERIA: A conceptual online tool to assess potential burn severity of wildfires in Mediterranean forests

Year: 2025

Type: article

Abstract: Fire plays an important role in the management of Mediterranean ecosystems but changes in fire regime influence the vulnerability and resilience of the plant-soil interphase. Concretely, burn severity is one of the most critical factors directly affecting plant community, soil characteristics and microbial activity. In this way, burn severity maps are crucial to provide information for policy makers, land managers and researchers to increase the extinction success and improve the adaptive forest management for fire prevention and post-fire restoration. Some approaches to generate burn or fire severity mapping have been developed using machine learning and artificial intelligence techniques, focused on fire danger analysis, elaborating comprehensive predictive models or using global data source after fire but our aim is to generate a digital tool to assess potential burn severity. The tool developed implements spatial results integrated in maps to predict potential burn severity based on a hierarchical architecture. We tailored the choice of algorithm for each step to the specific ecological dynamics of Mediterranean and Continental zones and validated the entire framework using robust group-by-fire cross-validation. We present SEVERIA as a free online tool to obtain potential locations of preventive treatments or support the design of urgent post-fire stabilization activities to mitigate potential negative effects and the restoration of burned areas. • We developed models to predict potential burn severity of wildfire. • Terrain-vegetation variables and climatic anomalies were integrated. • The proposed approach showed high accuracy for Mediterranean wildfires. • SEVERIA is a online tool providing open access sources as georeferenced layer.

Source: Ecological Informatics

DOI: 10.1016/j.ecoinf.2025.103496