Within the framework of the Karst Firewall 5.0 project, a comprehensive assessment of wildfire hazard and vulnerability has been carried out in the cross-border Karst/Carso area, with a particular focus on the impacts of current and future climate change. The deliverable provides a robust scientific basis to support more effective and coordinated cross-border wildfire management and contributes to the development of preventive and adaptive measures aimed at reducing wildfire risk.
The analysis combines two complementary methodological approaches: wildfire hazard probability modelling using the MaxEnt method and vulnerability assessment through Multi-Criteria Decision Analysis (MCDA) based on the Analytical Hierarchy Process (AHP). Both approaches rely on an extensive set of harmonised cross-border data, including anthropogenic, environmental, climatic and topographic factors, as well as on the analysis of more than 2,300 recorded wildfire ignition points.
The results show that anthropogenic factors, particularly proximity to transport infrastructure and land-use patterns, play a key role in wildfire occurrence and spread. The models demonstrate a high level of consistency between predicted areas of increased hazard and vulnerability and the locations of past wildfire events, confirming their reliability. Projections for the near future (2011–2040) indicate changes in the spatial distribution of wildfire hazard and an increase in areas characterised by very high vulnerability, further highlighting the need for a coordinated, cross-border and long-term approach to wildfire risk management.
The deliverable represents a significant scientific contribution to the understanding of wildfire dynamics in the Karst/Carso area and supports the development of effective strategies for wildfire prevention, preparedness, and response in the context of climate change.
Key findings
Wildfire hazard and vulnerability maps were developed for current and future climate conditions across the entire cross-border Karst/Carso area.
MaxEnt modelling achieved high statistical performance (average AUC = 0.754), demonstrating strong predictive capability for the spatial distribution of wildfire hazard.
Proximity to roads, railways and settlements is the most influential factor for both wildfire hazard and vulnerability, confirming the strong anthropogenic component.
More than 67% of recorded wildfires occurred within 50 m of road infrastructure.
Vulnerability analysis (MCDA/AHP) shows that over 90% of the area falls within high or very high vulnerability classes.
Future climate scenarios indicate an increase in areas of very high vulnerability, despite a potential reduction in hazard probability in some locations.
The results confirm that MaxEnt and MCDA/AHP are complementary approaches, together providing an integrated assessment of wildfire risk.
The deliverable provides a technical basis for strengthening cross-border cooperation, decision-making support and the development of wildfire risk reduction action plans.
