Geoscience Reference
In-Depth Information
types (Roy et al.
2006
). Other recent work has also been developing more automated
and regionally flexible systems for mapping severity using newer sensors such as
the Moderate Resolution Imaging Spectroradiometer (MODIS) (Loboda et al.
2007
;
Roy et al.
2005
).
Along with these developments in remote sensing, it has been necessary to
develop systematic protocols for assessing fire severity in the field. Key and Benson
have also introduced a measure they call the Composite Burn Index (CBI) that
is coordinated with a set of field techniques to provide a field system for assess-
ing vegetation, soil, and even longer term ecological effects of wildland fire (Key
and Benson
2006
). The CBI is also intended to provide a systematic method for
assessing fire severity in the field, that can be directly linked to the phenomena that
are being captured by the dNBR. Like the dNBI, the CBI has been developed for
the forests of western North America, and developmental forms of the CBI field
protocol have been used in the U.S. since 2001.
3.4.3 Timing
Once again our grassland and forest examples can provide starting points for think-
ing about the importance and complications of understanding the role of timing in
fire regimes. In essence we are interested in how often a given place burns as well
as in which seasons it tends to burn. In our grassland example we had historical evi-
dence that the area was subjected to fire every 2-6 years. Therefore we might refer
to this area having a Mean Fire Return Interval (MFRI) of 4 years. National data
sets are now available that map MFRI for the entire U.S.
However, like intensity and severity, the notion of MFRI becomes more complex
as we examine it more closely. One statistical problem is that knowing an area's
MFRI does not mean we understand the variability around that mean. A 4 year
mean might mask the possibility of much longer intervals between fires. It can also
remain unclear whether this fire regime is the result of human activities or is in fact a
“natural” regime. Additionally, it is unclear whether we expect the entire 1000 acre
area to burn every 4 years, or as is the case in many MFRI map products, we are
mapping the MFRI at exactly each pixel. For example Fig.
3.3
shows each pixel in
a region mapped for expected Mean Fire Return Interval (Fig.
3.3
).
One alternative to MFRI is to instead consider Fire Rotation Interval, or the
amount of time it would take for an entire region of interest to be visited by fire,
regardless of how many individual fires are required. Fire Rotation Interval may be
of more interest to many land managers responsible for specific management units,
but it is inherently subjective and tied to the size of the area of analysis, while MFRI
is a less subjective point measure.
In some cases it may be important to understand the historical seasonality of
the fire regime as well. Many organisms co-evolved with fire and have adapted
to historical fire seasonality. For example, many plants survive hot summer fires
by resprouting from their roots quickly after fire. However, prescribed fires often
take place when conditions are cooler and wetter. Paradoxically, the increased soil
