tend to be higher in freshwaters than in the oceans, and most freshwaters thus are
optically very complex. As noted earlier for chlorophyll, remote sensing methods
used for marine waters are not always applicable to freshwaters, and this situation
also applies to CDOM. For example, marine scientists interested in CDOM look
forward to using a band in the near UV (~380 nm) that will be available on a
forthcoming NASA sensor for ocean CDOM. Plant pigment absorbance decreases
greatly below ~400 nm, but CDOM absorbance continues to increase exponentially.
This band thus avoids interference between the absorbance of plant pigments and
CDOM that precludes using bands in the blue region for CDOM retrieval. Bands in
the near UV likely would not be useful for inland waters, however, because light
absorption by the typically higher levels of CDOM is so strong that there is
essentially no reflectance signal (all incoming light is absorbed).
Finally, one of the main reasons for measuring CDOM by ORS is the possibility
of using the values to estimate DOC in lakes at regional-to-global scales. As
discussed in Sect. 2.2.3, this is not straightforward because DOC-CDOM correlations are not always high. Even when they are, a relationship that works well for
one set of lakes may not be the same for a different set of lakes. A further
complication in DOC-CDOM relationships is the recent finding that complexation
of DOM by dissolved iron enhances the color intensity of the organic substances
[64, 65]. Scientists interested in using CDOM to estimate DOC at regional-toglobal scales should recognize that DOC-CDOM relationships are site specific and
perhaps time specific [29]. Additional predictor variables likely will be needed to
develop more robust predictive relationships between CDOM and DOC. In addition
to a possible need to account for the iron content of the water, water residence time
would help account for photobleaching of CDOM [66], the CDOM spectral slope
(S) would help define the quality or structural nature of CDOM [67, 68], and
various climatic and landscape metrics [69, 70] may account for DOC loadings to
lakes.
3 Current and Upcoming Remote Sensing Systems
for Regional Water Quality Assessment
A large number of airborne and space-borne sensors are potentially available for
remote sensing of water resources (Table 1), but none is ideally suited for monitoring inland waters, especially regarding our primary interest for this chapter—
water quality assessments of all lakes (above some nominal size) at regional scales.
Systems that are expensive, need to be tasked to collect specific imagery, cover only
small areas, or have coarse spatial resolution may be suitable for special projects
but not for routine synoptic lake monitoring. Moreover, sensors with only a few
broad bands do not provide reflectance data useful for accurate retrieval of water
quality measures like chlorophyll across a broad range of water quality conditions,
i.e., for optically complex inland waters. Characteristics of systems suitable for
regional aquatic assessments include spatial resolution appropriate for lakes > 4 ha
(i.e., spatial resolution or pixel size of 5–50 m
2 ), regular collection of imagery
126
L.G. Olmanson et al.
optically very complex. As noted earlier for chlorophyll, remote sensing methods
used for marine waters are not always applicable to freshwaters, and this situation
also applies to CDOM. For example, marine scientists interested in CDOM look
forward to using a band in the near UV (~380 nm) that will be available on a
forthcoming NASA sensor for ocean CDOM. Plant pigment absorbance decreases
greatly below ~400 nm, but CDOM absorbance continues to increase exponentially.
This band thus avoids interference between the absorbance of plant pigments and
CDOM that precludes using bands in the blue region for CDOM retrieval. Bands in
the near UV likely would not be useful for inland waters, however, because light
absorption by the typically higher levels of CDOM is so strong that there is
essentially no reflectance signal (all incoming light is absorbed).
Finally, one of the main reasons for measuring CDOM by ORS is the possibility
of using the values to estimate DOC in lakes at regional-to-global scales. As
discussed in Sect. 2.2.3, this is not straightforward because DOC-CDOM correlations are not always high. Even when they are, a relationship that works well for
one set of lakes may not be the same for a different set of lakes. A further
complication in DOC-CDOM relationships is the recent finding that complexation
of DOM by dissolved iron enhances the color intensity of the organic substances
[64, 65]. Scientists interested in using CDOM to estimate DOC at regional-toglobal scales should recognize that DOC-CDOM relationships are site specific and
perhaps time specific [29]. Additional predictor variables likely will be needed to
develop more robust predictive relationships between CDOM and DOC. In addition
to a possible need to account for the iron content of the water, water residence time
would help account for photobleaching of CDOM [66], the CDOM spectral slope
(S) would help define the quality or structural nature of CDOM [67, 68], and
various climatic and landscape metrics [69, 70] may account for DOC loadings to
lakes.
3 Current and Upcoming Remote Sensing Systems
for Regional Water Quality Assessment
A large number of airborne and space-borne sensors are potentially available for
remote sensing of water resources (Table 1), but none is ideally suited for monitoring inland waters, especially regarding our primary interest for this chapter—
water quality assessments of all lakes (above some nominal size) at regional scales.
Systems that are expensive, need to be tasked to collect specific imagery, cover only
small areas, or have coarse spatial resolution may be suitable for special projects
but not for routine synoptic lake monitoring. Moreover, sensors with only a few
broad bands do not provide reflectance data useful for accurate retrieval of water
quality measures like chlorophyll across a broad range of water quality conditions,
i.e., for optically complex inland waters. Characteristics of systems suitable for
regional aquatic assessments include spatial resolution appropriate for lakes > 4 ha
(i.e., spatial resolution or pixel size of 5–50 m
2 ), regular collection of imagery
126
L.G. Olmanson et al.