520
C.B. Field et al.
24.3.4 Vegetation Type
Large differences in the productive capacItIes of different ecosystem or
biome types (Whittaker and Likens 1975) makes an accurate map of vegetation distributions an important prerequisite for large-scale assessments of
photosynthesis and production. Most of the global-scale vegetation maps in
use today are based on surface observations (Matthews 1983) or climate
data (Holdridge 1947), but remote sensing is increasingly important. The
standard techniques for vegetation classification from satellite data require
knowledgeable individuals to identify training sites with known vegetation
types and then use statistical techniques, especially maximum likelihood, to
assign unknown areas (Richards and Kelly 1984). Mixture analysis, in which
the spectrum or time series of, for example, vegetation indices from each
pixel is deconvolved into a small set of endmember components, provides
another approach to land-surface classification. It offers several advantages,
of which one of the most promising for ecological studies is that pixels can
be treated as mixtures of components that mayor may not be traditional
ecosystem types (Ustin et al. 1993).
24.4 Remote Sensing of Physiological Status
24.4.1 Fluorescence
Chlorophyll fluorescence provides an extremely useful window on photosynthetic mechanisms, especially with recent improvements in the techniques for pulse-modulated measuring systems and analysis of fluorescence
lifetimes (Krause and Weis 1991). Almost all fluorescence measurements are
remote sensing, according to the definition we use here, but the diversity of
laboratory and field-oriented leaf-level techniques is beyond the scope of
this chapter.
In general, extending fluorescence studies from the leaf level to larger
scales has been difficult, because most techniques require controlled excitation. Extensions to larger scales are, however, beginning to look increasingly tractable. These extensions involve both passive techniques that assess
chlorophyll fluorescence under ambient conditions and active techniques
that provide controlled excitation. In marine ecosystems, ambient chlorophyll fluorescence creates increased upwelling in the red that can be used to
estimate chlorophyll concentration (Parslow and Harris 1990). Fluorescence
detection in the Fraunhofer lines, where atmospheric oxygen absorbs essentially all of the downwelling sunlight, provides a potentially useful approach
to eliminating interference from reflected light (Carter et al. 1990). Techniques for measuring laser-induced fluorescence (Chappelle and Williams
1987) are currently under development for canopy-scale and aircraft measurements (Lichtenthaler 1988).
C.B. Field et al.
24.3.4 Vegetation Type
Large differences in the productive capacItIes of different ecosystem or
biome types (Whittaker and Likens 1975) makes an accurate map of vegetation distributions an important prerequisite for large-scale assessments of
photosynthesis and production. Most of the global-scale vegetation maps in
use today are based on surface observations (Matthews 1983) or climate
data (Holdridge 1947), but remote sensing is increasingly important. The
standard techniques for vegetation classification from satellite data require
knowledgeable individuals to identify training sites with known vegetation
types and then use statistical techniques, especially maximum likelihood, to
assign unknown areas (Richards and Kelly 1984). Mixture analysis, in which
the spectrum or time series of, for example, vegetation indices from each
pixel is deconvolved into a small set of endmember components, provides
another approach to land-surface classification. It offers several advantages,
of which one of the most promising for ecological studies is that pixels can
be treated as mixtures of components that mayor may not be traditional
ecosystem types (Ustin et al. 1993).
24.4 Remote Sensing of Physiological Status
24.4.1 Fluorescence
Chlorophyll fluorescence provides an extremely useful window on photosynthetic mechanisms, especially with recent improvements in the techniques for pulse-modulated measuring systems and analysis of fluorescence
lifetimes (Krause and Weis 1991). Almost all fluorescence measurements are
remote sensing, according to the definition we use here, but the diversity of
laboratory and field-oriented leaf-level techniques is beyond the scope of
this chapter.
In general, extending fluorescence studies from the leaf level to larger
scales has been difficult, because most techniques require controlled excitation. Extensions to larger scales are, however, beginning to look increasingly tractable. These extensions involve both passive techniques that assess
chlorophyll fluorescence under ambient conditions and active techniques
that provide controlled excitation. In marine ecosystems, ambient chlorophyll fluorescence creates increased upwelling in the red that can be used to
estimate chlorophyll concentration (Parslow and Harris 1990). Fluorescence
detection in the Fraunhofer lines, where atmospheric oxygen absorbs essentially all of the downwelling sunlight, provides a potentially useful approach
to eliminating interference from reflected light (Carter et al. 1990). Techniques for measuring laser-induced fluorescence (Chappelle and Williams
1987) are currently under development for canopy-scale and aircraft measurements (Lichtenthaler 1988).