MSB-FRA: Scaling fire size from local process to continental pattern

Yale University, New Haven CT

Investigators

Abstract

Fire is a fundamental process in the Earth system, determining vegetation structure, changing atmospheric dynamics and carbon cycling, and damaging human infrastructure and livelihoods. One need only think of the recent, extreme fire season on the US West Coast for an example of how serious the impacts of fire can be. Predicting where fire will occur and how intense it will be is critical for both human and ecological systems in the future. Fire modules in global biosphere models allow forecasts of fire frequency and intensity. However, incorporating fire into global biosphere models efficiently and accurately remains difficult because global models run at a coarse scale relative to that at which many atmospheric, oceanic, and ecological processes operate. One of the key missing components is realistic fire extinctions. Currently, most global models force fires to go out after some pre-determined interval instead of allowing them to burn for days to months. The extreme fires contribute the majority of the impacts in the Earth system, and capturing their effects is critical to useful predictions. Improving the understanding and representation of fire extinction mechanisms in global fire models may improve predictions of fire occurrence in the Earth system. The broader impacts of this project will be improvements in fire management at field sites and savannas more broadly, and training of postdoctoral associates. An incomplete understanding of fire extinction forces global fire models to rely either on statistical representations of fire pattern or on local, unscaled assumptions about fire behavior. As a result, fire models diverge wildly, from each other and from reality, in their predictions for past, current, and future fire distributions in the Earth system. In this project, the researcher ask why fires go out, first in savanna ecosystems, where fires are frequent and therefore data plentiful, then expanding into other flammable systems (including coniferous and Mediterranean systems like those in Western North America). Such data feeds into theoretical generalizations of fire extinction (Phase 1). The team predicts that fire extent will result from interactions between random heterogeneity in fuels/landscapes and underlying topographic, weather, or anthropogenic patterns in the landscape. They will then implement the results of this work into global fire models, beginning with CLM4-DGVM and SPITFIRE-JSBACH and then proceeding more broadly via collaborations with the Fire Model Intercomparison Project (Phase 2). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.