Bernard Parresol, USDA Forest Service, Southern Research Station, (Presenter)


We measured fuel values from 629 plots systematically placed across the Savannah River Site, South Carolina, an 80267 ha land base. The site is ecologically diverse, with 35 Atlantic Coastal Plain forest types. A complete assessment was done of the surface fuels. In addition, canopy fuel parameters were measured. The site is composed of 6329 stands, and fuel values for each stand were statistically imputed from the inventory data. Thus populated with fuel values, we computed the fuel characteristic classification system (FCCS) fire behavior parameters for each stand, and performed hierarchical cluster analysis to group stands with similar fire behavior. From the clusters, centroid fuel values were computed. Stands close to the centroid values were used to build 7 custom fire behavior fuel models that were then calibrated to the FCCS parameters fire spread rate, flame length, and fireline intensity over a range of fuel moisture and wind speeds. Results indicate there is a range in fire behavior within forest types-age groups and deviation in fire behavior among the custom models. Geospatial fire models like FlamMap can then be used to produce a landscape representative of measured fuel conditions to help guide mitigation decisions.