Multispectral remote sensing of harmful algal blooms in Lake Champlain, USA.

ABSTRACT: This study developed satellite remote sensing models to detect cyanobacterial blooms via chlorophyll a in Lake Champlain. Landsat Enhanced Thematic Mapper Plus data was used to retrieve chlorophyll a concentrations, phytoplankton, and cyanobacteria biovolume by calibrating and validating with coincident observation data. Correlation analysis results showed that band 2 (green band) and the band ratio of 2/1 (green/blue) were most highly correlated to chlorophyll a concentration (r = 0.76 and 0.82, respectively). Multiple regression results identified band 2 and 3 (red), and band ratio of 2/1 and 3/1 (red/blue) as critical information to estimate chlorophyll a concentrations. The regression models accounted for 72 to 83% of the variability in chlorophyll a observations, allowing for estimates of phytoplankton and cyanobacteria levels in the lake. Satellite image processing results successfully showed the temporal and spatial distribution of chlorophyll a, phytoplankton, and cyanobacteria in the lake. This information can be used to evaluate the effect of pollution sources and weather conditions, and assist decision making for water management.

Assessment of nutrient distributions in Lake Champlain using satellite remote sensing.

The introduction of nutrients to lakes causing eutrophic conditions is a major problem around the world. Proper monitoring and modeling are important to effectively manage eutrophication in lake waters. The goal is to develop remote sensing models for nutrients, total phosphorus and total nitrogen, in Lake Champlain. The remote sensing models were created using multivariate linear regression with the unique band combinations of Landsat Enhanced Thematic Mapper Plus (ETM+) imagery based on the empirical relationship with the field observations. The resulting models successfully showed nutrient distributions in the most eutrophic part of Lake Champlain, Missisquoi Bay, with reasonable adjusted coefficient of determination values (R(2)=0.81 and 0.75 for total phosphorus and total nitrogen, respectively). The results show the feasibility and the utility of satellite imagery to detect spatial distributions of lake water quality constituents, which can be used to better understand nutrient distributions in Lake Champlain. This approach can be applicable to other lakes experiencing eutrophication assisting decision making when implementing Best Management Practices and other mitigation techniques to lakes.