Geoscience Reference
In-Depth Information
Fig. 3.4
FlamMap and FARSITE share the input data above. All input rasters must be co-
registered with a shared extent and pixel size. (From Finney
2006
)
private companies have been working for years to develop analytical models and
computer based modeling programs to help predict the spatial patterns of wildland
fires. All these projects must face the various complications discussed above as well
as a host of other technical and ecological issues.
Rather than produce an exhaustive list of the efforts underway, this chapter
focuses on just two major modeling packages that are now widely used in the U.S.
to model how fires will burn across complex landscapes. One is the PC computer
software package FARSITE, which is intended to help land managers and fire man-
agers predict the behavior and growth of active fires with known ignition points
and a known set of weather conditions. The other software package FlamMap, uses
similar input data sets and algorithms to predict expected fire behavior and fire
effects over an entire landscape regardless of ignition point. Rather than model-
ing the behavior and growth of a single fire, as with FARSITE, FlamMap predicts
the expected fire behavior for each pixel in a raster data set assuming fire visits each
pixel in a predetermined set of weather and fuel moisture conditions. Figure
3.4
shows an assortment of the input data layers necessary to run each of these pro-
grams (Fig.
3.4
). Information on both programs and several others is available from
www.firemodels.org
.
3.5.1 FARSITE, Fire Simulator
The software program FARSITE (Fire Area Simulator) has become a standard tool
for wildland fire modeling (Green et al.
1995
; Finney
1998
). It is widely used by the
National Park Service, Forest Service, and other federal and state land management
