Remote Sensing Inversion And Analysis Of The Water Transparency Of The Daihai Lake In The Past 35 Years

Lake is the main carrier of natural resources and an important part of the whole natural ecological environment system,which plays an important role in regulating regional climate,recording regional environmental evolution and maintaining the balance of regional ecosystem.Water transparency refers to the degree of water can make the light through,also known as the plug’s depth,is one of the important parameter of describing water body optics,is a key index of water quality of lakes,it and colored dissolved organic matter in water,suspended particulate matter and chlorophyll are closely related,can directly reflect the lake clear and cloudy,assess the degree of eutrophication of water bodies,Therefore,it has important scientific significance in the study of lake ecological environment.Remote sensing technology can be used to monitor the current situation and changing trend of lake water quality in real time and large area,which can effectively save manpower and material resources.Meanwhile,it can promote the theoretical and application value of remote sensing technology in monitoring and management of lake ecosystem.In this paper,based on Sentinel-2 MSI and Landsat-8 OLI images and taking Daihai Lake As the study area,the inversion model of water transparency was established.The accuracy of the established empirical model and BP neural network model was verified and applied,and the influence of meteorological factors and social factors on Daihai Lake transparency was analyzed.The results show that:（1）based on the regression analysis method,transparency experience model is established,including single band model,the band than the model and the hybrid band model,through the scatter diagram,correlation and error are compared,and the comprehensive results of mixed wave model inversion accuracy is best,determination coefficient R~2=0.63,the root mean square error（RMSE）of 0.25 m,The mean absolute percentage error（MAPE）was 21.07%.In the established BP neural network model,the determinant coefficient R~2of test set of MSI＿insitu＿27 model is 0.91,the root mean square error RMSE is 0.12m,and the average absolute percentage error MAPE is 21.69%.The BP neural network algorithm is better than empirical algorithm and semi-analytical algorithm.（2）BP neural network algorithm was applied to landsat-8 OLI and Sentinel-2 MSI images to obtain the temporal and spatial distribution characteristics of Daihai Lake water transparency over 35 years.The results show that the average transparency of Daihai Lake water varies from 0.48 m to 2.54m in 35 years,with the highest average transparency in 2011 and the lowest in 2020.The monthly mean diaphaneity ranges from 1.76 m to 2.59m.The mean diaphaneity in May is the lowest from April to October,and the mean diaphaneity in August is the highest.The spatial distribution of Daihai Lake transparency is higher in the middle,lower around,higher in the northwest and lower in the southeast.（3）Analysis of influencing factors of water transparency in Daihai Lake:meteorological factors and social factors.Meteorological factors mainly include monthly mean wind speed,monthly accumulated precipitation and monthly mean temperature.There is a significant negative correlation between the monthly mean wind speed and the average transparency in April-October from 1986 to 2020,and the correlation coefficient is-0.74.There is a positive correlation between the monthly average transparency and the monthly average accumulated precipitation,and the correlation coefficient is 0.87.There is a positive correlation between the monthly mean transparency and the monthly mean temperature,and the correlation coefficient is 0.77.Social factors mainly for liangcheng population,GDP and cultivated land area,the number of newborn piggy at the end of the other animals and all the gross value of industrial output was analyzed,and the results showed that adai seawater transparency and studied the correlation between the social factors of small,adai seawater transparency and liangcheng population phase relationship value is 0.23,and p value is 0.19>0.05;The correlation value between transparency and cultivated land area in Liangcheng County is 0.11,and the correlation value between transparency and year-end large livestock quantity in Liangcheng County is 0.27.The total industrial output value and gross domestic product of Liangcheng County are negatively correlated with transparency,the correlation coefficient is-0.07 and-0.24 respectively,and the P value is greater than 0.05,indicating that meteorological factors have a greater impact on the transparency of Daihai Lake water than social factors.