Remote Sensing Estimation Of Three-dimensional Green Volume And Gradient Characteristics Of Urban Green Space

The three-dimensional green volume inversion of urban green space and its gradient characteristics research can provide data basis and technical support for the rapid,accurate and comprehensive monitoring of urban green space.Taking the urban area of Xuzhou as an example,based on Sentinel-2A remote sensing data,field quadrat data,and DEM data,this thesis constructs and optimizes a three-dimensional inversion model of urban green space by using the stepwise multiple linear regression method by distinguishing the types of urban green space vegetation cover.,improve the accuracy of the model,construct the gradient distribution index from the three dimensions of three-dimensional green volume grade,pixel scale,and terrain scale,reveal the spatial gradient characteristics of the three-dimensional green volume of urban green space,and put forward suggestions for the management of urban green space in Xuzhou.The main research results and conclusions are as follows:(1)The three-dimensional green volume of urban quadratic green space was measured and the end-member information of urban green space vegetation was extracted.The three-dimensional green volume of urban green space in the quadratic data is estimated by the method of equation simulation,and its type structure is analyzed.Secondly,through the steps of minimum noise separation,endmember primary selection,and purification,the vegetation endmembers of urban green space are obtained,and the mixed pixels are decomposed by the least square method to obtain the vegetation coverage abundance map of urban green space.The root mean square of the decomposition is 0.012.It shows that the quality of the mixed pixel decomposition results is good.(2)Build and optimize the three-dimensional green volume inversion model of urban green space.Based on the imaging process of remote sensing images,the analysis of the ecological structure and function of urban green space,and the experience of predecessors,66 variables from five levels of original band,terrain,vegetation index,texture,and vegetation abundance were selected,and an inversion factor set was constructed.The multiple regression method constructs a three-dimensional green volume inversion model of urban green space.Then,the inversion model was optimized on the basis of distinguishing urban green space cover,and the model goodness-of-fit(R2)value for accuracy verification increased from 0.7173 to 0.8432.(3)The spatial gradient characteristics of the three-dimensional green volume of urban green space are revealed.The optimized inversion model was used to invert the three-dimensional green volume of urban green space in the study area,and the gradient characteristics of the three-dimensional green volume of urban green space were analyzed from the three dimensions of three-dimensional green volume grade scale,pixel scale,and terrain dimension.The study found that the gradient areas with high altitude and high slope are concentrated and the vertical space filling is rich,and the gradient areas with low altitude and low slope are scattered and the vertical space filling is sparse.In addition,the horizontal coverage and vertical distribution of the threedimensional green volume of urban green space show strong spatial differences.The thesis has 33 figures,13 tables,and 170 references.