Mapping future fire probability under climate change: Does vegetation matter?

Area of Science:

• Ecology
• Environmental Science
• Geospatial Analysis

Background:

• Fire pattern prediction is crucial for effective management and policy.
• Statistical fire models often omit vegetation data, introducing uncertainty.
• Understanding vegetation's role in fire dynamics is key to improving projections.

Purpose of the Study:

• To assess the sensitivity of future fire projections to the inclusion of static and dynamic vegetation data.
• To compare fire models with and without vegetation variables under different climate scenarios.
• To evaluate the impact of vegetation changes, including wildfire effects, on fire probability.

Main Methods:

• Developed a statistical fire modeling framework incorporating various vegetation maps.
• Compared models with no vegetation, static vegetation, and dynamic vegetation scenarios (fire and land use change).
• Projected future fire probability for all and large fires (>40 ha) in the Sierra Nevada, California, under two climate scenarios.

Main Results:

• Fire projections showed high sensitivity to the inclusion and type of vegetation data used.
• Models incorporating wildfire's effect on vegetation predicted lower future fire probability.
• Some scenarios yielded opposing trends in future fire extent and probability, highlighting significant uncertainty.

Conclusions:

• Vegetation variables are critical predictors in fire models, and their omission introduces substantial uncertainty.
• Accounting for vegetation dynamics, especially wildfire impacts, is necessary for more realistic fire projections.
• Integrating statistical and process-based models offers a promising approach to capture a wider range of fire scenarios.