323 / 2019-02-20 21:27:47

Analysis of PM2.5 Spatial Distribution Pattern at Urban Scale Based on Multi-source Data ------a Case Study of Wuhan

PM2.5,AOD,SARA,GBDT,Urban scale,Wuhan

全体主题 > Remote Sensing

PM2.5 pollution has been a crucial problem affecting the environment and urban sustainable development. However, urban scale PM2.5 spatial distribution pattern didn’t receive too much attention yet. In this study, based on multi-source data, we used Gradient Boost Decision Tree (GBDT) with powerful fitting capacity to obtain continuous PM2.5 concentration distribution in autumn and winter of Wuhan. In these multi-source data, Aerosol Optical Depth (AOD) retrieved by Simplified high resolution MODIS Aerosol Retrieval Algorithm (SARA) ensured the resolution of the result to be 500m which can satisfy the urban scale research, and R-square (R^2) of the model was well performed reaching to 0.83. Then we adopted hotspot analysis to find the areas of high PM2.5 concentration and analyzed the relationship between PM2.5 concentration and industry, residential life and traffic in these areas. It shows that there is serious PM2.5 pollution around the third ring road and different areas have different reasons for high PM2.5 concentration.