Modeling the climate change impact on wildfire

To explicitly estimate how climate change may affect insured losses from wildfire, it is necessary to connect the expected changes in climate to expected changes in fire frequency and burned area.

There are various approaches for making the connection between climate change and fire activity. One approach uses statistical models to develop relationships between the climate and burned area using historical information, and then apply these relationships to expected future changes in climate, under the assumption that the historically based relationships do not change. The future changes in climate are simulated from GCMs. GCMs provide the most advanced simulations of future climate change, using physics to represent processes in the atmosphere, ocean, land, and sea ice, to study how the climate may respond to changing atmospheric levels of greenhouse gases, most notably carbon dioxide (CO₂).

Another approach is to obtain information about how burned area and other wildfire characteristics may change because of future climate from coupled climate fire models (CCFMs). In addition to the basic components that they share with GCMs, CCFMs also have dynamic global vegetation models (DGVMs) and physics-based fire models. The fire models run interactively – using information about vegetation (e.g., biomass and type) and climate (VPD, precipitation, wind, and more) – and can simulate fires that ignite at locations and frequencies based on historical data and simulated weather conditions. The fire models can feed the burned area information back to the atmospheric component of the CCFM and thus can influence subsequent weather and climate simulation as well as future wildfire risk.

Additionally, catastrophe models are used in conjunction with one or both of the two approaches summarized above to create a stochastic view of how the risk will change.