An algorithm for the retrieval of suspended sediment in coastal waters of China from AVHRR data

An algorithm using an analytical model based on the difference of the NOAA/AVHRR (Advanced Very High Resolution Radiometer) Channel 1 (580-680 nm) and Channel 2 (720-1100 nm) reflectance data is developed for the retrieval of suspended sediment in coastal and shelf waters. The model assumes that the suspended sediment concentration, S, is a function of the difference of the water leaving reflectance of Channels 1 and 2, R(d). The formula is governed by the optical parameters of water and suspended sediment, including volume scattering and absorption coefficients of the two channels. The analytical model yielded a best fit curve when the water leaving reflectance of the two channels were plotted against each other for Case 2 water, where the suspended sediment concentration ranged from 5 to over 100 g m^-3. A standard curve of S against R(d) was obtained. Using Tassan's (1994) recommendations for the contribution of water and suspended sediment to the in-water absorption and backscattering coefficients in his three-component color model, the suspended sediment concentration corresponding to the maximum point of R(d) was about 60 g m^-3, and was independent of the atmospheric optical properties. Thus, this process provides a convenient tool to remove the atmospheric fluctuations of atmospheric transmittance by reconciling the maximum point of the image with the standard curve. The algorithm was tested using data from seven transects in the China Sea, and the retrieved results for the Zhujiang (Pearl River) Estuary were compared with the sea-truth data with good agreement. This suggests that the algorithm can be used as a seasonal regional model for water masses along the China coast.