Publication Date
2-2012
Abstract
We developed empirical remote sensing models to estimate chlorophyll
a concentrations and
cyanobacteria synoptically, over a large inland lake using available Landsat Enhanced Thematic
Mapper Plus (ETM+) data. In contrast to previous studies which rely on the spectral
characteristics of the cyanobacterial specific pigment, phycocyanin, we developed remote
sensing models capable of directly detecting cyanobacterial biovolume. This distinction is
important because Landsat ETM+ data lacks the spectral band information required for optimal
phycocyanin detection. Each model was calibrated and cross-validated with existing
in situ
measurements
from Lake Champlain’s Long-Term Water Quality and Biological Monitoring
Program (LTMP). Lake station measurements taken between 2006 and 2009 were matched with
radiometrically converted exoatmospheric reflectance data from seven spectral bands on the
Landsat ETM+ sensor. Step-wise multi-linear regression indicated data from Landsat ETM+
bands 1, 2 and 3 were most significant for predicting chl-
a and cyanobacteria biovolume. Based
on statistical analysis, the linear models that included visible band ratios slightly outperformed
single band models. The final models captured the extents of cyanobacterial blooms throughout
the 2006-2009 study period. The results serve as an added monitoring tool for resource
managers and present new insight into the initiation and propagation of cyanobacterial blooms in
Lake Champlain.