Geoscience Reference
In-Depth Information
within a particular fire regime. Often our historical information is incomplete and we
really know very little about the sizes or internal heterogeneity of fires that occurred
before written record keeping. In areas with particularly long MFRIs spanning cen-
turies or even millennia, it can be very difficult to ascertain whether the sizes of a
set of fires fall outside the natural range of variability.
In some ecosystems fire patterns remain visible in the landscape for centuries.
This is particularly true in higher elevation forests with extremely long MFRIs and
forest types that burn in large continuous blocks. As the forests regenerate, going
through successional stages, these previous burns can be obvious hundreds of years
later, even in satellite imagery. In other areas, the patterns left on the ground by fire
are far less obvious, but in some of these areas Dendrochronology can once again
be a valuable approach to understanding this element of fire regimes. For example,
landscape level tree sampling can allow researchers to identify years in which many
trees where scarred by fire. By mapping these trees and dates on a landscape it is
sometimes possible to extrapolate the likely size of historical fires (Morgan et al.
2001
).
Mapping the internal heterogeneity of historical fires is even more difficult
than mapping perimeters. It can even be very difficult to map the internal patterns
of last season's fires, much less historical fires. This is made even more difficult
by the continuous nature of fire effects. Fires do not burn in a binary way across
landscapes. At almost any scale of analysis, or pixel size, it is common to have
many pixels that contain a mix of both burned and unburned vegetation. Therefore,
most fire maps grossly underestimate the true complexity of the patterns on the
ground. To make things even more difficult these patterns may express themselves
differentially over time after a fire. For example, trees that seem healthy immedi-
ately after a fire can succumb later to processes such as insect infestations that were
exacerbated by the fire.
Because of complications like those discussed above, fire regimes can seem
impossibly complex and variable. Nevertheless the concept of fire regimes can pro-
vide considerable guidance in how we think about fires in any given area. In the
same way that complicated variability in fuels can be simplified into the 13 stan-
dard fire behavior fuel models, fire regimes are often simplified into a few simple
categories. For example, the national LANDFIRE data sets discussed at the end of
this chapter has a data set that maps the fire regimes of the entire U.S. into just the
five categories seen in Fig.
3.1
. Depending on the intended use, such coarse catego-
rization may be acceptable, while for other uses extremely detailed local fire regime
information may be necessary.
3.5 Predictive Fire Modeling and Data
There are many exciting developments occurring in fire mapping, geospatial
research in fire ecology, and wildland fire management. One of the most engaging of
these geotechnical developments is predicative computer modeling of wildland fire
behavior and fire effects. Many government agencies, academic researchers, and
