Print

Wheeler, Sarah M.

Natural Resources

2006

MS

Harmful algal blooms in Lake Champlain (VT-NY-QC) are of increasing concern to scientists and the public. Such blooms impact drinking water supplies, recreational use of lake waters, and, if toxic, pose a threat to human health. An improved means of mapping and monitoring the distribution of such blooms over current point sampling methods is needed. Remote sensing offers promise but current operational sensors are limited in either their spectral or spatial resolution for freshwater applications. This study assessed the utility of three operational satellite sensors to detect and quantify cyanobacteria blooms in Lake Champlain. Data from QuickBird, SPOT, and MERIS multispectral sensors were acquired coincident with in situ water samples in the summers of 2003 and 2004. In-vivo fluorescence of chlorophyll a and phycocyanin, and transmissivity were measured along 5-15 km transects located with GPS. Models were developed to relate satellite radiances (as single, band, band ratio, or principal component measures) to pigment concentrations, used as a proxy for cyanobacteria density. For SPOT and QuickBird, simple linear regression was used to relate imagery with pigments and the coefficient of determination (r²) value was used, to compare models. The increased spectral information provided with MERIS imagery allowed for the application of semi-empirical optical models.