Most remote sensing studies of water composition have made use, not of the a priori deductive method outlined above, but of an empirical approach in which relations have been sought between remotely measured radiances in particular wavebands, or functions of one or more radiances or reflectances, and the concentrations of specific components such as phytoplankton and suspended solids determined by in situ measurements within water bodies. Once such relations are established it should in principle be possible to use them to determine the optical properties and composition of water bodies from remote sensing data alone. This is an active field of research, which has given rise to a very large quantity of published work in recent years. The treatment here will not be encyclopaedic: rather, general principles are outlined and a variety of representative examples are discussed.
A useful concept in this context is that of a retrieval variable,1341 which is any function of the reflectance (or radiance) values in one or more wavebands, X[R(lj)], that is empirically found to be related to the concentration of some component of the water by a simple algorithm. Thus, the problem in remote sensing is usually to identify a suitable X, and to develop a simple equation, by means of which the concentration of the constituent of interest may be calculated from X.
Was this article helpful?