Research On Water Body Extraction Based On Remote Sensing Images And Analysis Of Dynamic Changes In The Poyang Lake Are

Water is the source of life and the foundation of all things,which are the basis of human survival and development.In recent years,due to natural and man-made factors,freshwater resources are in short supply in various parts of China.To ensure the sustainability and stability of water resources,it is of great practical importance to measure and grasp the changes of water bodies in real time.Based on landsat data,this paper analyzes the spectral characteristics of various types of features in the images and constructs a water index model;The model is used to extract the area of Poyang Lake in the high and low water periods from 2003 to 2021,and analyze the dynamic changes;The response relationship between the interannual change of Poyang Lake water area and the impact factors in the past 18 years is analyzed by combining the seven impact factors,including runoff,precipitation and evaporation.The main research contents include:(1)Based on landsat images,the optical reflection characteristics of water and other surface features were studied.It was found that subtracting the DN values of the near-infrared,SWIR1,and SWIR2 bands from the sum of DN values in the green and red light bands can effectively separate water.Therefore,a Five-band Comprehensive Water Index(FBCWI)was constructed.Five different scenarios were selected and nine commonly used water index models were used for the extraction experiments.The results show that the overall accuracy of the model is greater than 96.99%,the kappa coefficient is greater than 93.98%and the total error is less than 5.97%,which can not only effectively remove common noise,but also extract fine water.Compared with other indices,the five-band integrated water index model constructed in this paper has significantly improved the extraction accuracy.(2)The area of Poyang Lake from 2003 to 2021 is extracted using Landsat-5,Landsat-7 and Landsat-8 image data,and its dynamic changes are analyzed.The extracted area data were also divided into three time periods:2003-2008,2010-2015and 2016-2021,and the mean,standard deviation and other statistics were calculated for change characteristics analysis.The results show that:from 2003 to 2021,the area of Poyang Lake in the wet season and dry season showed a zigzag change.The maximum area in the wet season and dry season was 4054.13 km~2 and 2360.94 km~2respectively in 2020,and the minimum area was 2021.31 km~2 and 1350.48 km~2respectively in 2013 and 2014;During 2003-2021,the annual area of Poyang Lake fluctuated around 3300km~2 in wet season and 1800 km~2 in dry season;The standard deviation of the three periods during the high water period gradually increases,with an increase of about 50%per period,and is relatively stable during the low water period.(3)According to the change of Poyang Lake area from 2003 to 2021,combined with the data of runoff,precipitation,evaporation,average temperature and land type,the influencing factors of Poyang Lake area change are analyzed.The results show that the change of Poyang Lake area has a strong positive correlation with runoff and precipitation,and a strong negative correlation with evaporation in both wet and dry seasons.In terms of wet season,the correlation coefficients between the area change of Poyang Lake and runoff,precipitation and evaporation are 0.808,0.771 and-0.688respectively;During the dry season,the correlation coefficients are 0.783,0.700,and-0.734,respectively;The correlation coefficient between the area of Poyang Lake in wet season and dry season and the average temperature,farmland area,forest area and impervious surface area is less than 0.2,almost no correlation.In addition,multiple regression equations are established for runoff,precipitation,evaporation and Poyang Lake area change in wet season and dry season respectively.The results showed that the R~2 of the regression model during the wet season was 0.675,the correlation coefficient R was 0.822,and the probability of significance was 0.005;The R~2 of the dry season model is 0.652,the correlation coefficient R is 0.807,and the probability of significance is 0.007.It shows that the regression equation has a good fitting effect,which can explain the reasons for the change of Poyang Lake area to a certain extent.