Remote Sensing Estimation And Spatiotemporal Evolution Analysis Of Urban Surface Runoff In Xuzhou Based On Impervious Surface

Urbanization is an inevitable result of the evolution of social and economic development.With the rapid advancement of urbanization,urban impervious surfaces have gradually replaced natural landscapes such as vegetation and bare soil as typical urban surface types.It improves the confluence rate and flow rate of urban surface runoff,changes the urban natural drainage system,and increases the risk of urban waterlogging.Therefore,it is necessary to study the relationship between urban surface runoff and impervious surface.This thesis uses the four Landsat images of Xuzhou in1997,2005,2013 and 2020 as the basic data.Using the method of combining remote sensing with GIS,qualitative and quantitative,based on the extraction of impervious surface information,remote sensing estimation and spatiotemporal evolution of surface runoff in Xuzhou city are carried out.The following are the main research findings and conclusions:(1)Modified linear spectral mixture decomposition model for impervious surface information extraction.In order to effectively solve the serious problem of underestimating the impervious surface of the village due to the confusion between the impervious surface and the bare soil in the traditional linear spectral mixture decomposition model.From the two stages of endmember purification and unmixing post-processing,the linear spectral mixture decomposition model for the extraction of impervious surface information is improved.The errors of the results after endmember purification are all lower than those of the unpurified results in the RMS accuracy test.In 1997,the RMS accuracy are 0.0094,0.0150 in 2005,0.0092 in 2013,and 0.0159 in2020.Compared in the RMSE error and R~2 test,the RMSE error results of the traditional LSMA model,the endmember purification results and the improved LSMA model are:0.2618,0.2226 and 0.1079 respectively,and the R~2 test results are 0.545,0.671 and 0.923 respectively;In the comparison of the results over the same period,the impervious surface in spring and summer in 2020,the mean abundances are 0.3591 and0.3336,and the extraction results are very close in numerical and spatial distribution,which proves the reliability of the improved LSMA model to extract urban impervious surfaces.(2)Estimation of surface runoff in different rainfall scenarios in Xuzhou urban area.After using the improved SCS-CN model to calibrate the runoff curve parameters CN value,initial loss coefficientλand other core parameters.According to the rainfall grade standard of the China Meteorological Administration and the 10-year rainfall total in Xuzhou City,determine the 24-hour cumulative rainfall of heavy rain,heavy rain,heavy rain and extra-heavy rain,and the total rainfall in dry years and wet years,so as to realize remote sensing of surface runoff in each scenario.The results show that the surface runoff in Xuzhou urban area shows an increasing trend year by year from1997 to 2020.The main urban area is the main high-value area of runoff,and the surrounding vegetation area is relatively less.In the indirect accuracy verification,the average runoff of the 50 typical flood-prone points in the main urban area of Xuzhou in 2020 reached 58.98 mm,which is much higher than the average runoff of 28.21 mm in the urban area of Xuzhou,reflecting the high accuracy of remote sensing estimation of surface runoff.(3)Revealing the spatiotemporal evolution characteristics of urban surface runoff in Xuzhou.In the research on the grade response of surface runoff,it is found that the grade response of surface runoff is more frequent with the increase of the impervious surface of the city,and the grade response of urban surface runoff is more obvious when the rainfall is small.The median center is shifted in the east-west direction as a whole and relatively concentrated.With the increase of rainfall,the north-south offset increases,and the data is more discrete.The research on the evolution degree of annual surface runoff shows that the sensitivity of urban surface runoff attitude will decrease with the increase of rainfall,and the dynamic degree of the wet year scenario is smaller than that of the dry year scenario;the research on the evolution and distribution of annual surface runoff shows that the overall direction of the runoff data in 2013 and2020 is stronger and more concentrated,and the flattening is lower in the surface runoff grade.When the fluctuation is large and the level is high,it tends to be close.(4)Delineated the risk pattern of urban waterlogging in various regions of Xuzhou.The risk area categories are divided into four categories:high risk area,potential risk area,medium risk area,low risk area,and the difference between level II and level I.The results show that the spatial heterogeneity of the urban waterlogging risk pattern in Xuzhou is obvious.The high-risk areas are mainly located in the main urban area of Xuzhou,Jiawang District and some townships in Tongshan District.The high-risk area of Grade II covers an area of 302.57km~2,accounting for 10.17%.The low-risk areas are mainly located in large tracts of farmland and other vegetation areas,with a maximum area of 1283.03 km~2,accounting for 43.13%.In the regional analysis,the high-risk areas in Quanshan District and Gulou District have a relatively high proportion,while the Tongshan District has the largest area of high-risk areas and low-risk areas due to the largest administrative area,but the low-risk areas account for 48.48%and the high-risk areas account for a small proportion.The high-risk areas in the delineation pattern should be the key areas for urban flood control and waterlogging prevention in Xuzhou.There are 48 figures,31 tables and 173 references in this thesis.