Application of Multi-spectral Satellite Imagery in Monitoring of Aquatic Vegetation and Water Quality Parameters in Large Inland Waters

Abstract

Following the great potential of optical remote sensing and its increased application in quality assessment of inland waters, we have developed time dependent vegetation abundance prediction models based on its statistical relationship with the concentrations of total suspended matter (TSM) and phytoplankton chlorophyll (Chl-a) water quality (WQ) parameters in the lake as well as the amount of rainfall in its drainage basin. We start by retrieving the selected WQ parameters from MERIS (Medium Resolution Imaging Spectrometer) multispectral satellite imagery of Lake Victoria based on their optical properties, and obtain their seasonal variations over the period 2003 to 2010. We then carry out regression analysis to establish the time dependent statistical correlation between estimated vegetation abundance and the retrieved WQ constituents as well as rainfall after various response periods, and identify an optimal response period for each precursor