Assessment of Spatial Distribution of Algal Biomass in a Subtropical Lake in China by Modeling of Landsat ETM+ Data Coupled with Field Data

Abstract

The objective of this research is to construct an efficient way of monitoring water quality and assessing trophic state using remote sensing techniques in Qinshan Lake of Hangzhou, China. Two Landsat ETM+ images were acquired and simultaneous in situ measurements, sampling and analysis were conducted. Results of the study indicated that the ratio of ETM+1/ETM+3 was the most effective single band in estimating chlorophyll-a (Chl-a), followed by normalized ratio vegetation index (NRVI). Two multiple regression models with determination coefficients were further constructed between logarithmically transformed Chl-a and the combination of ETM+1/ETM+3, ETM+2/ETM+3, and ETM+3/ETM+4 of ten sample sites. The resulting models, Log (Chl-a)=1.65 + 0.87*(ETM+1 / ETM+3) 3.39*(ETM+2 / ETM+3) + 0.89*(ETM+3 / ETM+4), and Log (Chl-a)=2.94 1.37*(ETM+1 / ETM+3) + 0.40*(ETM+2 / ETM+3) 0.20*(ETM+3 / ETM+4), both showed strong ability to evaluate the distribution of Chl-a, with R2 of 0.72 and 0.92, respectively. Then two trophic state maps generated for Qinshan Lake using this model could identify zones with a higher potential for eutrophication, which turned out to be an appropriate method for synoptic monitoring of water quality in lakes. Similar modeling can be made for any given subtropical lake, to provide rapid and long term assessment of water quality and also useful information for decision making.