Geoscience Reference
In-Depth Information
this section we discuss the contribution of fluorescence spectroscopy to the validation
of CDOM remote sensing algorithms, and comment on the use of Light Detection and
Ranging (LIDAR) and passive measurements of natural CDOM fluorescence.
6.7.1 Fluorescent CDOM and Validation of Remote Sensing Products
Flow-through CDOM fluorometers are useful to validate remote sensing retrievals of a g .
Traditionally, remote sensing data are compared to field data collected within ±3 hours of
a satellite overpass (Bailey and Werdell, 2006 ). Often the comparison is between one field
value and a remote sensing retrieval that represents an average over the size of the pixel
(from 250 m 2 to ~1 km 2 or more depending on sensor, viewing angle, and spatial binning
used). This comparison assumes low spatial and temporal variability and may be correct
in the open ocean. However, the assumption can be incorrect in coastal environments with
high spatial variability (Yuan et al., 2005 ). For example, a perfect satellite retrieval of a g
(representing the average a g in a pixel) could be different from any number of discrete sam-
ples collected within that pixel. This may give the impression of inaccurate remote sensing
retrievals. Collecting a large number of discrete samples within the area of a typical pixel
for analysis in a bench-top fluorometer or spectrophotometer is not a practical solution.
However, flow-through CDOM fluorometers can be well calibrated using discrete samples
to produce accurate estimates of CDOM absorption coefficient at wavelengths relevant
to remote sensing (i.e., Ferrari and Tassan, 1991 ; Hoge et al., 1993 ; Green and Blough,
1994 ; Del Castillo et al., 1999 ). For example, Hoge et al. ( 1993 ) demonstrated the use of
fluorescent CDOM (FCDOM) to retrieve values of a g , and later applied this technique to
the calibration of airborne LIDAR to retrieve a g values (Hoge et al., 1995 ). Their linear
regression curves between a g and FCDOM had r 2 values between 0.89 and 0.98, but typi-
cally >0.98. Similar results have been reported elsewhere (Blough et al., 1993 ; Del Castillo
et al., 1999 ).
Because the fluorescence spectrum of CDOM is broad, and its absorption spectrum
monotonic, it is easy to establish relationships between FCDOM and a g over a large range
of wavelengths. Nevertheless, for the purpose of using this technique to validate remote
sensing retrievals of CDOM, it is convenient to report values at wavelengths covered
by satellite sensors, and to provide data on the absorption spectra used to calibrate the
FCDOM to a g including spectral slope S and how it was calculated. This facilitates com-
parison between published data.
The user must be aware of changes in S (Blough and Del Vecchio, 2002 ) and fluores-
cence efficiency ( ϕ f ) (Blough and Del Vecchio, 2002 and references therein; Lakowicz,
2006 ) along salinity gradients. A constant S allows for reconstruction of a CDOM absorp-
tion spectrum based on measurement of a g at any λ (within the λ range used to calcu-
late S ). A constant ϕ f indicates that CDOM fluorescence is proportional to the amount
of light absorbed by CDOM, allowing to calculate a g ( λ ) from FCDOM. These are typi-
cally observed at salinities >30, when the marine CDOM end-member starts to influence
Search WWH ::




Custom Search