Projection of wildfire activity in southern California in the mid-twenty-first century

Abstract

We estimate area burned in southern California at mid-century (2046-2065) for the Intergovernmental Panel on Climate Change (IPCC) A1B scenario. We develop both regressions and a parameterization to predict area burned in three ecoregions, and apply present-day (1981-2000) and future meteorology from the suite of general circulation models (GCMs) to these fire prediction tools. The regressions account for the impacts of both current and antecedent meteorological factors on wildfire activity and explain 40-46% of the variance in area burned during 1980-2009. The parameterization yields area burned as a function of temperature, precipitation, and relative humidity, and includes the impact of Santa Ana wind and other geographical factors on wildfires. It explains 38% of the variance in area burned over southern California as a whole, and 64% of the variance in southwestern California. The parameterization also captures the seasonality of wildfires in three ecoregions of southern California. Using the regressions, we find that area burned likely doubles in Southwestern California by midcentury, and increases by 35% in the Sierra Nevada and 10% in central western California. The parameterization suggests a likely increase of 40% in area burned in southwestern California and 50% in the Sierra Nevada by midcentury. It also predicts a longer fire season in southwestern California due to warmer and drier conditions on Santa Ana days in November. Our method provides robust estimates of area burned at midcentury, a key metric which can be used to calculate the fire-related effects on air quality, human health, and the associated costs.