Over the past 25 years wildfires have caused more than $10 billion (in today’s dollars) in insured property losses in the United States, with the majority of losses occurring in California. The 1991 Oakland Hills Fire destroyed 2,900 structures and cost insurers more than $3.5 billion. The 2003 Cedar and Old Fires in Southern California destroyed 3,700 homes and cost insurers more than $2 billion. In the "October Fire Siege" of 2007, 23 wildfires raged simultaneously, ultimately causing more than $2 billion in insured losses and the destruction of some 3,300 homes and other structures; nearly a million people were displaced and nine lost their lives. 

Increasing Wildfire Losses

Losses from wildfires have been increasing. Urban expansion continues to penetrate into previously undeveloped forests, chaparral, and grasslands—areas at high risk of wildfire. Nearly 40 percent of new home development in the Western United States is occurring in this "wildland-urban interface" (WUI), putting more insured property at risk every year. In California alone, it is estimated that more than five million homes are located in this ecological habitat.

AIR’s wildfire model for California helps insurers, reinsurers, and intermediaries manage wildfire risk for both individual policies and entire portfolios of properties. A fully probabilistic model, the model allows insurers to better identify their exposures relative to areas at risk and to estimate their potential losses.

High-Resolution Modeling

The AIR model incorporates at high resolution the factors that drive wildfire ignition, growth (in size and intensity) and spread—as well as the factors that contribute to or mitigate insured losses, such as the construction characteristics of potentially vulnerable structures. The model makes use of the "LANDFIRE" fuels database, the most comprehensive and up-to-date source of information on vegetation—its types, locations, densities, volatilities—currently available.

Fire intensity is modeled at a 30-meter resolution, and the model’s fire-spread algorithm takes into account topography, local wind speeds, and fire suppression activities. An Urban Conflagration Effect (UCE) explicitly captures the possibility that a WUI fire will be transformed into an urban conflagration, such as the fire that occurred in Oakland Hills in 1991.

AIR currently offers a wildfire model for: