This research focussed on understanding the relationship between water quality parameters, the inherent optical properties, and the subsurface irradiance reflectance, for application in remote sensing. The multitemporal and multi-sensor applicability of the derived algorithms has been demonstrated by the successful application to remote sensing data from two instruments, flown on two days under different conditions. The strength of this approach lies in the ability for combining information and results from underwater light climate research as well as in situ and remote sensing reflectance measurements.

Improvements are possible if more information concerning the scattering and backscattering (and, therefore, volume scattering functions) of the phytoplankton and the tripton was available. Therefore a decrease in analytical and an increase in semi-empirical elements in the algorithms were required, in successive order, for chlorophyll a, cyanophycocyanin (CP-cyanin), seston dry weight (DW), vertical attenuation coefficient for downwelling irradiance (Kd), and Secchi disk transparency (SD). Chlorophyll a, CP-cyanin and seston are parameters that influence the optical characteristics of water, whereas Kd and SD are a function of the optical properties of the water.

The large range in optical water quality parameters in the study area has led to new insights: the variability in specific absorption and scattering coefficients within and between the four different water types indicates that the development of remote sensing algorithms for inland water quality analysis must reckon with adjustable parameterisation of absorption and scattering variables for each water type studied.

The subsurface irradiance reflectance, R(0-), is the essential parameter for applications of the analytical model for remote sensing. It is essential because R(0-) can be determined from the inherent optical properties of a absorption coef.), b (scattering coef.) and B(q) (volume-scattering function, where q = angle of scattering), from in situ spectroradiometric measurements and from remote sensing measurements.

For chlorophyll a and CP-cyanin concentration estimation the analytical model provided the means to perform a sensitivity analysis on the influence of other parameters such as aquatic humus and tripton. Natural variations in aquatic humus and tripton levels in these waters led to 35 ug 1^-1 and 10 ug 1^-1 chlorophyll a equivalent error respectively. This indicates that estimates of the other inherent optical properties are required for successful application of these algorithms.

From the combined absorption and scattering coefficients of all parameters it was deduced that for remote sensing the spectral areas where only one dominant spectral feature is present will be the most useful for determining inland water quality. Such features include the 624 nm CP-cyanin and the 676 nm chlorophyll a absorption features and the overall absorption minimum at 700 - 710 nm. Below 500 nm the effects of absorption by aquatic humus, phytoplankton and tripton and the high level of scattering are compounded, presenting a four-parameter equation of influence on the reflectance signal measured by remote sensing.