THE RELATIONSHIP BETWEEN SPECTRAL REFLECTANCE, ABSORPTION AND
BACKSCATTERING FOR FOUR INLAND WATER TYPES
Dekker 1 , A.G., Hoogenboom 1 , H.J., Goddijn 1 , L.M. & Malthus 2 , T.J.M.
‘Inst, for Environmental Studies, Vrije Universiteit, De Boelelaan 1115, 1081 HV Amsterdam, The Netherlands.
2 Sch. of Applied Sciences, University of Wolverhampton, Wulfrunastreet, Wolverhampton, WV1 1SB, UK.
ABSTRACT
The relationship between spectral reflectance, absorption and backscattering is essential for the development of analyti
cal and multitemporal algorithms for remote sensing of inland waters. The spectral subsurface irradiance reflectance
R(O-), absorption a , scattering b and backscattering b b were determined for 18 waters covering four inland water types.
Physical models relating R(O-) to the inherent optical properties were investigated. The linear backscatter albedo model
R(O-) = r, b b / (a + b b ) is the most suitable model with r, ranging from 0.12 to 0.56 with an average of 0.29, with
a relative error of 1 % to 5 %. The variation in r, between the water bodies was large. In contrast to the other three
water types the eutrophic lakes showed a spectral dependence of r, with R(O-) . The main recommendation is that mea
surements of the volume scattering functions of the various fractions of the particulate material in inland waters are
required.
KEY WORDS: water quality, optical modelling, spectral reflectance, inherent optical properties.
1. INTRODUCTION
The range of optical water quality properties that may be estimated by remote sensing has increased from suspended
matter to cyanophycocyanin and chlorophyll a contents, vertical attenuation coefficients, transparency and aquatic humus
(Dekker, 1992 & 1993). This progress in the development of algorithms for remote sensing of inland waters is due to
a change from an empirical to an analytical approach, producing quantitative maps of water quality, and, to the increas
ed spectral and radiometric resolution of sensors. The analytical approach is preferable since it allows sensitivity
analysis and, hence, further improvement of the accuracy of the water quality parameter estimation.
In the analytical approach physical relations are derived between the water quality properties, the underwater
light field and remote sensing measurements. The underwater tight field is determined by the inherent optical properties
(IOP) of the water constituents, i.c. the absorption coefficient a and the volume scattering function P(0) from which
the scattering (b) and backscattering (bj coefficients may be determined. The IOP are physically related to the
subsurface irradiance reflectance R(O-) which is a key parameter linking the IOP to the (ir)radiance measurements from
remote sensing or in situ data. R(O-) is only slightly dependent on solar elevation, atmospheric or water surface con
ditions and is therefore called a quasi-inherent property. Algorithms developed on the basis of R(O-) , irrespective of
the mode of calculation or measurement, have multitemporal validity.
The IOP and the R(O-) for 18 inland water bodies in The Netherlands were measured, representing four inland
water types. Each water type has its own characteristic algae and suspended matter composition resulting in varying
magnitudes of the IOP. The four types are: 1) shallow eutrophic lakes characterized by high phytoplankton concentra
tions (mainly filamentous prokaryotes) and tripton concentrations (both detrims and resuspended bottom material);
2 ) shallow mesotrophic lakes characterized by low phytoplankton concentrations but relatively high tripton levels;
3) deep lakes with variable phytoplankton concentrations dominated by spherical cyanobacteria and tripton (detrims)
and 4) river and canal waters characterized by relatively low phytoplankton concentrations but high tripton levels.
245
