Based On LANDSAT 8 And HJ-1A/B CCD Remote Sensing Image Of Dongting Lake Water Quality Conditions Of The Study

With the development of society,the impact of water environment on human beings is increasingly strengthened.Traditional water quality monitoring methods need to set up on-site monitoring stations,and the monitoring scope is difficult to represent the whole area of water,time-consuming and laborious,the effect is not satisfactory,so a new monitoring model is needed to monitor the Dongting Lake area.In this paper,the water quality of Dongting Lake is studied and analyzed.The main research work is as follows:(1)Since LANDSAT series satellite images cannot completely cover the whole Dongting Lake area,this paper uses LANDSAT images and HJ-1A/B CCD environmental small satellite data to study and analyze the study area.After radiometric calibration,atmospheric correction,geometric correction and other image pretreatment,The preliminary image results of the study area were obtained.Landsat 8 and HJ-1A/B CCD environmental satellite were used as data sources to extract the water area of Dongting Lake.Three water area extraction models,namely single-band threshold method,normalized water index method and normalized vegetation index method,were used to compare the three water area extraction models by random sampling method.The error rates of the three models were-22.6%,-13.5% and-3.5%respectively,indicating that normalized vegetation index had the best effect in extracting water area.(2)Through the selection of pure pixels at the vertex of the two-dimensional scatter plot,it can be judged that the eutrophication of Dongting Lake occurred in the Lu Lake in the southwest of East Dongting Lake,which was in the mild eutrophication state,while the other lakes were in the moderate eutrophication state.The linear mixed pixel decomposition model was used to inversion the eutrophication state of Dongting Lake,and the state trend was consistent with the eutrophication result of two-dimensional scatter vertex end element.The spatial-temporal analysis of water quality and eutrophication in Dongting Lake in August 2013,July 2016,November 2019 and November 2021 shows that the eutrophication has improved greatly from mild eutrophication in 2013 to moderate eutrophication in 2021,while the water quality has fluctuated from category 3 to Category 4.(3)the use of BP neural network to a single site of 9 characteristic factors of water quality on 110 groups of data for training analysis,through the threshold value and the limitation of the number of iterations,water quality forecast data by BP neural network prediction error less than 5%,has high credibility,it can be predicted a priori basis data for future water governance,Good suggestions are provided to the government for water treatment.Combined with the result of monitoring and forecasting model,found that the water quality of Dongting lake area problem more severe east Dongting lake area is mainly the southwest part of the lake,the lake water quality is relatively good,therefore,the government needs to strengthen the regulation of the lake area of water quality monitoring,to control eutrophication index,submitted to the corresponding water quality survey report,ensure the quality of the water quality of Dongting lake area.