Agriculture Reference
In-Depth Information
ridgelines that then become turbulent and erratic. These erratic winds converge at
ridgetops and can contribute to long-distance spotting (Pyne
et al
.
1996
).
Barriers of reduced fuel loading, which could include rocks, rivers, alluvial fans
and other terrain features, may inhibit fire spread. This topographic effect varies
through the season; as drought progresses, different portions of the landscape are
added as available fuels, contributing to a temporal component of landscape fuel
heterogeneity. Steep terrain also enhances long-distance spread of firebrands.
Fire Behavior and Landscape Models
Models have played a fundamental role in fire science and fire management. Of
widest use are fire behavior models that predict fire activity based on equations
that relate fire spread to surface fuel characteristics and environmental factors
(Rothermel
1972
). Many models have been developed but two of the more widely
used are BEHAVE (Andrews
1986
) and FARSITE (Finney
1998
). The former
predicts fire characteristics associated with different categories of fuel and envir-
onmental data and the latter gives spatially explicit predictions of fire spread.
One of the major limitations to modeling fire behavior is the lack of detailed fuel
bed data for most landscapes. As a consequence there has been extensive study of
fuel characteristics for a wide range of vegetation types and these have been
formulated into fuel models that can be used in fire spread models (Scott &
Burgan
2005
). In recent years there has been a proliferation of fire behavior
models designed to overcome certain limitations in previous models (see review
in Peterson
et al
.
2007
).
One of the primary limitations of fire behavior models is that they often are based
on physical models of fire spread in dry surface fuels and are not easily adapted to
dealing with crown fires in live and dead shrublands. Alternative approaches to this
“first principles” approach are probabilistic models of fire behavior based on
empirically determined relationships between shrubland fuels, environmental par-
ameters and observed fire characteristics (e.g. Bilgili & Saglam
2003
).
Fire spread models have found a wide range of uses. Managers often use these
to predict fire spread during fire events in order to more effectively deploy fire
suppression crews. Both managers and researchers use these models to evaluate
fuel management impacts on subsequent fire behavior (e.g. van Wagtendonk
1996
;
Graham
et al.
1999
).
Percolation models have proved useful in understanding fire behavior and
impacts on vegetation in crown fire ecosystems by testing the influence of fuels
and weather on fire spread between adjacent cells (Turner & Romme
1994
). Such
models have been applied to questions of prescription burning impacts (Bradstock
et al.
1998a
) or to test assumptions about the role of fuel age mosaics in controlling
the size of wildfires (Zedler & Seiger
2000
; Keeley & Zedler
2009
).
Models such as FIRESCAPE, LANDSUM or FATELAND have been
developed to investigate landscape patterns of burning under different fire regimes
