Simulating the role of fire in forest structure and functional type coexistence: Testing FATES-SPITFIRE in California forests

Earth system models are representing increasing levels of complexity in the processes that shape the structure and distribution of vegetation types on the landscape to better capture changes driven by climate warming and land use. The western U.S., and California in particular, have seen recent increases in multiple aspects of fire, and projections suggest that climate change will further accentuate these trends. Climate, fire and vegetation structure are intimately linked, requiring that they be examined and simulated together. The new vegetation demographic model FATES has been coupled to the SPITFIRE fire behavior and effects model to represent the role of fire in shaping biome boundaries and forest structure. FATES explicitly simulates tree recruitment, growth, and mortality from underlying physiological processes that are responsive to environmental conditions, while SPITFIRE is dependent on vegetation state and environmental conditions, and predicts the effects of fire on tree mortality. Therefore, this model framework is suited to simulating interactions between global change, vegetation and fire at large scales. While FATES-SPITFIRE has been tested and applied across the savannah-forest ecotone in the tropics, its application in extratropical ecosystems is still emerging.

We aim to use FATES-SPITFIRE to examine how climate change, together with historical and future shifts in fire occurrence, affect the distribution and structure of vegetation in California. Here we focus on reproducing the historical behavior and effects of fire in mid-elevation mixed conifer forests of the Sierra Nevada mountains. To achieve this, we have adapted SPITFIRE to utilize spatially and temporally varying lightning ignition probabilities or, alternately, total (natural and anthropogenic) actual ignitions for the historic period (1980-2016). We compare FATES-SPITFIRE simulations versus available observations of burned area and burn severity, and quantify the impacts of fire on fuel consumption, tree mortality across size classes, and pre- versus post-fire functional type composition. Fire is a process integral to the historic and future structure of vegetation in California. Projections of vegetation change require accurate simulation of climate-fire-vegetation interactions.