Spatially explicit fire behavior models require measurements of elevation, slope, aspect, weather and vegetation to simulate fire behavior across the landscape. The vegetation parameters, collectively termed fuel models, used in these simulations vary based upon the specific requirements of the model chosen to forecast fire spread. In general, most fire behavior models will require estimates of the following variables in order to generate predictions: total aboveground biomass contributing to the forward rate of spread of the fire; mean stand height; fuel particle surface-area-to-volume (SAV); and moisture content.
Acquiring these estimates can be very costly and, as a result, the use of spatial fuels datasets mapped exclusively from field inventory data is generally limited to small-scale projects with an area on the order of 0.5 km2 or less. Spatial fuels data for areas exceeding this size are typically derived from one of the thirteen standard National Forest Fire Laboratory (NFFL) fuel models described in Anderson's 1982 publication, "Aids to Determining Fuel Models For Estimating Fire Behavior".
The NFFL fuel models assist in generating spatial fuels data by providing suggested reasonable values for each fuel property based upon the stand species composition and site condition. Existing spatial data in a Geographic Information System (GIS), such as vegetation and fire history, is oftentimes used to assign the most appropriate fuel model to each area in a heuristic fashion. In many cases, rather than simply relying solely on the thirteen NFFL models, additional more customized fuel models that may or may not have been developed from field inventory, are also used.