Regression Of Chlorophyll Content Based On Hyperspectral Remote Sensing Data In Chaohu Lake

With the rapid economic development and the intensification of industrialization and the impact of human activities, lake pollution and eutrophication problems have become increasingly serious. It is necessary to enhance lake water quality monitoring and treatment. But it is time-consuming for routine water quality monitoring, and it is difficult to reflect the lake’s overall water quality conditions. Remote sensing technology can quickly and timely provision of water quality in the lake, with the special advantage of monitoring in a wide range of space, low cost and long-term dynamic and facilitate monitoring, so it has great application in monitoring lake water quality.This paper summarizes and analyzes the water quality of the remote sensing monitoring at home and abroad, the theory, technology and method for Chaohu lake area, through the measured spectrum measurement Chaohu lake water, this paper analyzes the optic reflection characteristics Chaohu lake water, we adopt half experience and three band half analysis method to establish the remote sensing hyperspectral images of chlorophyll a quantitative estimation model, and found that, three band model ratio model and the difference compared with higher precision of the model inversion, correlation coefficient reached0.889. RMS error for12.39ug/L; And land statistics on three band model optimization, optimized the RMS error of inversion model fell2.28ug/L, correlation coefficient increases by3.9%.Construction and application of the model through the following steps to achieve:the characteristics analysis based on measured water spectrum, to find out the value of peaks and valleys, and the band which largest correlate with chlorophyll, and with the concentration of chlorophyll to establish multiple regression model, through regression modeling inversion to study the area of spatial distribution of chlorophyll concentration. Then use Kriging to correct the residuals of the concentration of chlorophyll in regression model, that reduce the error of the regression model inversion.The results measured by the test and found that the analysis of inversion model has a certain accuracy, which can effectively characterize Chaohu lake spatial distribution of chlorophyll concentration.