Accuracy Verification And Analysis On Water Reflectance Of Inland Lake

Inland water is the principle part of urban water resource and the monitoring of theinland lakes environment is necessary steps in protecting the quality of urban waterresource. With the environment problem of lake ecological increasingly prominent,using satellite remote sensing to monitor the lake water has become more and moreurgent. Remote sensing technology has the characteristics of rapid, widespread andcyclical, it is much better than conventional water quality monitoring in lake waterquality monitoring, and also is the future development trend of the inland water qualitymonitoring means. This paper take DaLiNuoEr lake, taihu lake and qinghai lake as thestudy area, using landsat5TM data and domestic satellite HJ1-A/B CCD multi-spectraldata, getting real reflectivity by inversion, and validation and analysis the inversionresults. Specific content as follows:1、Data pre-processing. Including the data processing of water spectral collectedin field, radiation correction, mosaic and tailoring of HJ image and TM image,projection transformation, splicing, cutting and pixel value transition of MODIS surfacereflectance products and standard aerosol products, AERONET aerosol data processing,etc.2、Atmospheric correction. This paper used the6S model method based on theradiative transfer model, get the real reflectivity of water. For HJ data, using a methodbased on lookup table to correct the atmospheric effects; While for TM images, use themethod directly based on6s radiative transfer.3、Accuracy validation and analysis of water reflectivity. Use water reflectivitydata collected in field and MODIS surface reflectance products to validate the inversionresults of HJ data and TM data, then analysis the error about it. According to thereflectivity inversion results of satellite data water, comparative analysis the watermonitor quality effect about the two kinds of sensors, HJ and TM, and compare thethree inland lake water body components.Based on the above analysis, the following conclusions are gained: the landsatellite multispectral sensor is also suitable for inland water bodies detection; TM images have less noise than HJ image in monitor the second water lakes; the inversioneffect of visible light band is better than that of infrared wave band; Composition oflake water clarity, qinghai lake is the highest, DaLiNuoEr lake is second, taihu lake islowest.