Fire and the Fire Regime Framework - Fire in Mediterranean Ecosystems - page 35

on steep slopes this relationship is not always reliable (Weise & Biging 1996 ;
Keeley 2009 ). Since fireline intensity is determined by both rate of spread and
fuels consumed, it is possible for two very different fires to have similar fireline
intensities: for example, low heat output with a high rate of spread and high heat
output with a low rate of spread (Pyne et al. 1996 ).
Fireline intensity is an established metric in forested ecosystems as there is a
well-documented relationship between flame length and scorching height of tree
crowns. However, fireline intensity often cannot explain mortality patterns since
mortality may be more a function of total heat output reflected in flame residence
time or a function of smoldering combustion in the duff after the flame front
passes (Wade 1993 ). Many other fire effects are not well predicted by fireline
intensity. Soil duff consumption, for example, is more related to temperatures at
the soil surface and the duration of heating (Ryan & Frandsen 1991 ; Miyanishi
2001 ). Also, survival of seedbanks or rhizomes may be more closely tied to
duration of heating as well as maximum soil temperatures than to fireline intensity
(Beadle 1940 ; Flinn & Wein 1977 ; Auld & O'Connell 1991 ; Bradstock & Auld
1995 ). This should come as no surprise since often very little radiant or convective
heat from combustion of aerial fuels is transferred to the soil, and generally soil
temperatures are more dependent on consumption of fine fuels on the surface
(Bradstock & Auld 1995 ). Although fireline intensity provides information for fire
fighters concerned with fire containment, resource managers may be more con-
cerned with temperature and duration of heating (residence time) as these may be
critical to retention of sensitive ecosystem components. In the future, fire man-
agers will likely depend heavily on remote imaging technologies for fire intensity
and these do not always scale with fireline intensity (Smith et al. 2005 ).
Due to the difficulties of measuring fire intensity, particularly for unplanned
wildfires, the terms fire severity or burn severity have been used as a postfire
indicator of intensity. Some definitions of fire severity have been rather general
statements about broad impacts of fires, such as the degree of environmental
change caused by fire (e.g. White & Pickett 1985 ; Simard 1991 ; Jain et al. 2004 ),
and consequently have not lent themselves to operationally useful metrics. Most
empirical studies that have measured fire severity have had a common basis that
centers on the loss or destruction of aboveground or belowground organic matter
(Keeley 2009 ). In forests, height of bole scorch or crown volume scorch are two
common measures and these correlate with fire intensity (Cheney 1981 ; McCaw
et al. 1997 ; Catchpole 2000 ). In shrublands, fire severity is commonly assessed
using the twig diameter on standing skeletons, which correlates with a crude
measure of fire intensity (Moreno & Oechel 1989 ;P ´ rez and Moreno 1998 ) and
similar measures have been used in heathlands (Whight & Bradstock 1999 ).
On landscapes prone to large wildfires, assessing fire severity has benefited from
the use of Landsat satellite imagery, which generates indices correlated with
ground measures of fire severity (Conard et al. 2002 ; Miller & Yool 2002 ; Chafer
et al. 2004 ; Hammill & Bradstock 2006 ; Keeley et al. 2008 ; Veraverbeke et al.
2010 ). Indices based on the Normalized Differenced Vegetation Index (NDVI)