Design of a Fire Susceptibility Index for fire risk monitoring

Wildfire risk assessment has traditionally been carried out using models based on meteorological parameters measured at spatially sparse weather stations. Remote sensing techniques can offer a cost effective way of esimating the required parameters such as fuel moisture and fuel temperature in near-real time. Moreover remote sensing can provide wide area coverage and thus circumvent errors introduced due to spatial interpolations of weather station data. In this paper we present a new remote sensing based Fire Susceptibility Index (FSI) based on the susceptibility of the underlying live fuel to burn. The proposed index is based on the concept of heat energy of pre-ignition and thus allows a physical meaning to be associated to the index values. Heat energy of pre-ignition is the heat energy required to bring a fuel from its current temperature to ignition temperature and can be estimated as the heat required in evaporating the moisture content plus the heat required to raise the temperature of the dry fuel to ignition temperature. The computation of the index requires inputs of fuel temperature and fuel moisture content, both of which can be estimated using remote sensing techniques. While MODIS surface temperature is used as a proxy for fuel temperature, fuel moisture is estimated by a linear regression technique utilizing the correlation between ground observations and the ratio of normalized difference vegetation index and surface temperature. Results are shown for the Georgia region during the spring and summer months of 2004.