511
Soil and Vegetation Moisture Changes for Forest Fire Events
the near-infrared (NIR) and visible red bands, the NDVI measures the changes in
chlorophyll content and in spongy mesophyll within the vegetation canopy. Several
variations of NDVI have been triggered thereafter for drought monitoring such as
the vegetation condition index (Kogan 1990, 1995), the anomaly of NDVI (Anyamba
et al. 2001), and the standardized vegetation index (Peters et al. 2002). However, the
major limitation of NDVI for drought monitoring is the apparent time lag between
a rainfall deficit and NDVI response (Reed 1993; Di et al. 1994; Rundquist and
Harrington 2000; Wang et al. 2001), since it provides information on vegetation
greenness (chlorophyll), which is not directly and uniformly related to the quantity
of water in the vegetation (Ceccato et al. 2002).
Water stress causes physiological changes in vegetation, which in turn causes
changes in the vegetation spectral signature (Marshall and Zhou 2004). Thus, drought
monitoring is accomplished by observing the spectral variations of water absorption
characteristics in the NIR and shortwave infrared (SWIR) regions (Marshall and
Zhou 2004). The SWIR reflectance reflects changes in both the vegetation water
content and the spongy mesophyll structure in vegetation canopies, while the NIR
reflectance is affected by the leaf internal structure and leaf dry matter content but
not by water content. Gao (1996) produced the normalized difference water index
(NDWI) using the SWIR channel as the water absorption–sensitive band and NIR
channel as the insensitive band. The combination of the NIR with the SWIR removes
variations induced by the leaf internal structure and leaf dry matter content, improving the accuracy in retrieving the vegetation water content (Ceccato et al. 2001).
While the first generation of remote sensing–based drought indices, either NDVI
or NDWI, relied on the few optical bands as provided by traditional sensors such as
NOAA AVHRR or Landsat TM, a new generation attempts to make use of the multiband capabilities of, for example, the MODIS sensor onboard the Terra and Aqua
satellites. A recently developed index is the NMDI as proposed by Wang and Qu
(2007). The NMDI is based on one NIR and two SWIR channels, exploiting the slope
of the two water-sensitive absorption bands 6 and 7 of MODIS. It is essentially an
improvement of the traditional normalized indices that usually use one sensitive band
and one insensitive band. Section 21.3 will review the theoretical basis of NMDI.
For reliable and operational use of satellite remote sensing–based indices in
drought monitoring, it is necessary to compare with ground measurements or traditionally used meteorological drought indices so as to establish how well the remotely
derived indices reflect wet and dry conditions. Sections 21.4 and 21.5 attempted to
evaluate the ability of NMDI for measuring and monitoring the drought conditions by
comparing with the meteorological data, to investigate the relation between drought
and wildfire occurrence, and to estimate the performance of NMDI for detecting forest fires burning in southern Georgia, USA, and southern Greece in 2007.
21.3 NMDI: A NORMALIZED MULTIBAND DROUGHT INDEX
FOR MONITORING SOIL AND VEGETATION MOISTURE
In the applications of spectral variation of water absorption bands, several indices using
reflectances from the NIR and SWIR channels have been proposed for remote sensing of vegetation water content from space. The most popular NIR–SWIR vegetation
Soil and Vegetation Moisture Changes for Forest Fire Events
the near-infrared (NIR) and visible red bands, the NDVI measures the changes in
chlorophyll content and in spongy mesophyll within the vegetation canopy. Several
variations of NDVI have been triggered thereafter for drought monitoring such as
the vegetation condition index (Kogan 1990, 1995), the anomaly of NDVI (Anyamba
et al. 2001), and the standardized vegetation index (Peters et al. 2002). However, the
major limitation of NDVI for drought monitoring is the apparent time lag between
a rainfall deficit and NDVI response (Reed 1993; Di et al. 1994; Rundquist and
Harrington 2000; Wang et al. 2001), since it provides information on vegetation
greenness (chlorophyll), which is not directly and uniformly related to the quantity
of water in the vegetation (Ceccato et al. 2002).
Water stress causes physiological changes in vegetation, which in turn causes
changes in the vegetation spectral signature (Marshall and Zhou 2004). Thus, drought
monitoring is accomplished by observing the spectral variations of water absorption
characteristics in the NIR and shortwave infrared (SWIR) regions (Marshall and
Zhou 2004). The SWIR reflectance reflects changes in both the vegetation water
content and the spongy mesophyll structure in vegetation canopies, while the NIR
reflectance is affected by the leaf internal structure and leaf dry matter content but
not by water content. Gao (1996) produced the normalized difference water index
(NDWI) using the SWIR channel as the water absorption–sensitive band and NIR
channel as the insensitive band. The combination of the NIR with the SWIR removes
variations induced by the leaf internal structure and leaf dry matter content, improving the accuracy in retrieving the vegetation water content (Ceccato et al. 2001).
While the first generation of remote sensing–based drought indices, either NDVI
or NDWI, relied on the few optical bands as provided by traditional sensors such as
NOAA AVHRR or Landsat TM, a new generation attempts to make use of the multiband capabilities of, for example, the MODIS sensor onboard the Terra and Aqua
satellites. A recently developed index is the NMDI as proposed by Wang and Qu
(2007). The NMDI is based on one NIR and two SWIR channels, exploiting the slope
of the two water-sensitive absorption bands 6 and 7 of MODIS. It is essentially an
improvement of the traditional normalized indices that usually use one sensitive band
and one insensitive band. Section 21.3 will review the theoretical basis of NMDI.
For reliable and operational use of satellite remote sensing–based indices in
drought monitoring, it is necessary to compare with ground measurements or traditionally used meteorological drought indices so as to establish how well the remotely
derived indices reflect wet and dry conditions. Sections 21.4 and 21.5 attempted to
evaluate the ability of NMDI for measuring and monitoring the drought conditions by
comparing with the meteorological data, to investigate the relation between drought
and wildfire occurrence, and to estimate the performance of NMDI for detecting forest fires burning in southern Georgia, USA, and southern Greece in 2007.
21.3 NMDI: A NORMALIZED MULTIBAND DROUGHT INDEX
FOR MONITORING SOIL AND VEGETATION MOISTURE
In the applications of spectral variation of water absorption bands, several indices using
reflectances from the NIR and SWIR channels have been proposed for remote sensing of vegetation water content from space. The most popular NIR–SWIR vegetation