Remote Sensing Invervsion Of Water Quality Parameters And Trohpic State Assessment In Poyang Lake

Water is the blood of the ecosystem,is the most important and most dynamic factors in global environment.However,the impact of with rapid economic development and industrialization and other human activities, increasing failure of freshwater resources, water pollution and eutrophication,especially in lakes, rivers.The conventional method of water quality monitoring in lakes which are time and money-consumed can't provide water quality information of the whole lake. Based on remote sensing technology to monitor water quality has great potential to get a wide range,fast,low cost and ease of dynamic monitoring of long-term advantages in the lake water quality monitoring.For the study area of Poyang Lake, field-based water quality sampling, measurement spectrometer, MODIS image three type data of simultaneous. Measured spectrum showed that water body presentsⅡcommon feature in Poyang Lake, chlorophyll a concentrations were lower, the peak was not obvious, obvious peak of suspended solids. After atmospheric correction of MODIS images based on image feature selection dark pixel method of Chavez, Between the corrected single-band, the band combined reflectance and chlorophyll a concentration, suspended solids concentration to do peason correlation analysis,showed that:there has been a lot of bands and band combinations with chlorophyll a presents highly related, the best resolution of the inversion band is 500m (B2+B7); suspension has the characteristics of time and space, only a small number of band or band combination present a significant correlation,the best inversion band is 500m resolution (B2-B7).Based on the results of MODIS correlation analysis, regression and BP neural network were used to establish inverse model. Model test results showed:500m resolution with strong radiation data is better data than 250m; when sufficient sample, BP network was used to build non-linear mapping model, relative to conventional statistical regression modeling shows a stronger ability of nonlinear mapping.Finally,between the measured concentration of water quality parameters do correlation analysis, results showed that between suspended matter and the transparency have significant correlation,and get mathematical expression, with Carlson trophic state index and the modified trophic state index formula, get TLI index model based on remote sensing technology. Combination of MODIS images, application model generation the the TSI index distribution of Poyang Lake.