Introducing SCENFIRE, a Post-Processing of Scenario-Based for the integration of Wildfire simulations

Wildfire risk assessment has evolved significantly over decades of research, enhancing both the accuracy and practical application of methods to evaluate wildfire occurrences and impacts. However, global changes and climate variability present challenges to quantitative assessments due to the diversity of future scenarios. In this context, we introduce SCENFIRE, a specialized selection algorithm designed to align simulated fire perimeters with specific fire size distribution scenarios.

The foundation of this approach lies in generating a vast collection of plausible simulated fires across a wide range of conditions, assuming a random pattern of ignition. The algorithm then assembles individual fire perimeters based on their specific probabilities of occurrence, determined by (i) the likelihood of ignition and (ii) the probability of particular fire-weather scenarios, including wind speed and direction.

This method offers several significant advantages. First, it eliminates the need for fine-tuning simulation parameters by creating an extensive pool of scenarios, which can be automated using scripting tools such as FConstMTT batch processing. Second, it allows for easy adaptation to various fire size distributions without necessitating recalibration of the simulation process.

The approach is exemplified in the eastern Mediterranean coast of Spain, a region prone to wildfires due to natural conditions and land abandonment. This area has experienced recurring large fire events over recent decades, making it an ideal setting to demonstrate the method's effectiveness.