Abstract:

A study case for retrieving trophic level index of water body of lakes in Wuhan，Hubei Province，using CBERS2 images data was presented.The aim of this study is to evaluate the probability of CBERS2 images data on estimating the eutrophic level of inland water.At first step，the “real” TLI (both aggregative trophic level index (TLI) and revision carlson index (TSIM)) at the observation spots of some lakes in Wuhan were calculated using the monitoring water quality data.In the mean time，the mean grey values of a 9×9 pixels sampling window for band1band4 at the corresponding spots were obtained and calculated from the CBERS2 image which had been taken radiometric calibration and geometric correction beforehand.After that，an experience inversion model for TLI had been built through multivariate regression using the mean grey values of each band as independent variables.Finally，the TLI of all water bodies were inversed with this model and a map for its spatial distribution was drawn up.Results demonstrate that there are good correlation relationships between the natural logarithm values of TLI and spectral reflectance of CBERS2 image，the square regression coefficients (R2）is about 0.51.The distribution of inversed TLI of the lakes fixes the reality quite well.This might provide us a cheap way for monitoring inland water quality since CBERS2 images data can be obtained freely from many data delivering centers in China.

Key words: trophic level index/ remote sensing/lake/ CBERS2