Study On Extraction And Dynamic Change Of Urban Forest Based On RS

The development of urbanization has triggered a series of ecological and environmental problems including noise pollution,solid waste pollution,urban water pollution and shortage,urban heat island effect and air pollution.Vegetable in urban space has important ecological value.It can improve urban ecological environment,mitigate urban climate effect,improve urban environmental quality and protect biodiversity through photosynthesis of vegetation.Therefore,it is of great significance to study the change of vegetation area in cities with rapid development.This paper selects six municipal districts in Beijing as the study area for the study of urban forest change.Using Landsat TM/OLI remote sensing image as data source and based on the spectral characteristic curve of the objects in the study area,this paper uses 11 vegetation indices,such as TNDVI and RVI,to process and analyze the data,and selects a vegetation index which can separate vegetation from background objects.Then it is used to process the images of different periods in the study area.Subsequently,the changes of urban forest coverage in the districts,municipal districts and ring roads in Beijing are analyzed and studied.The results are as follows:(1)Comparing the vegetation extraction results of 11 vegetation indices,it is found that the vegetation and plant can be separated highly by GEMI,NDGI and GVI.The GEMI's separability of vegetation and unused land is poor,but NDGI and GVI can classify vegetation and other land features.In actual classification extraction,NDGI needs to select two thresholds,while GVI only needs one threshold,which reduces the error sources of classification results.Finally,GVI is selected as the vegetation index of vegetation extraction experiment.(2)From 1992 to 2017,the urban forest coverage in Beijing research area showed a trend of decreasing first and then increasing.The urban forest coverage decreased from 42.94% in 1992 to 29.9% in 1994,and then increased to 39.09% in 2017.In the six municipal districts,the urban forest coverage of Dongcheng District and Xicheng District showed a continuous increasing trend.The urban forest coverage of Chaoyang District,Fengtai District and Shijingshan District showed the same trend as that of the whole study area.The urban forest coverage of Haidian District decreased during 1992-1999,and then increased continuously.(3)Based on the correlation analysis between the economic and population statistics of Beijing and the urban forest coverage in the corresponding areas.The results show that there is a strong positive correlation between per capita disposable income and urban forest coverage in Dongcheng District and Xicheng District,where urban forest coverage increases and population density decreases continuously,with the correlation coefficients of 0.999 and 0.992,respectively.There was a strong negative correlation between population density and urban forest coverage,and the correlation coefficients were-0.927 and-0.871,respectively in Dongcheng District and Xicheng District.In other municipalities,the correlation is weak.