Abstract
We integrated hyperspectral and field-measured chlorophyll-a (Chl-a) data from the Kristalbad constructed wetland in the Netherlands. We developed a best-fit band ratio empirical algorithm to generate a distribution map of Chl-a concentration (C
chla) from SPOT 6 imagery. The C
chla retrieved from remote sensing was compared with a water quality model established for a wetland pond system. The retrieved satellite results were combined with a water quality model to simulate and predict the changes in phytoplankton levels. The regression model provides good retrievals for Chl-a. The imagery-derived C
chla performed well in calibrating the simulation results. For each pond, the modeled C
chla showed a range of values similar to the Chl-a data derived from SPOT 6 imagery (10–25 mg m−3). The imagery-derived and prediction model results could be used as the guiding analytical tools to provide information covering an entire study area and to inform policies.
Development of Bayesian Monte Carlo techniques for water quality model uncertainty