Spatial Coupling Between Urban Heat Island Effect And AOD Distribution And Their Influencing Factors In Wuhan,china

Since the 21st century,the world has experienced rapid urbanization,which has led to urban environmental problems such as heat island effect(UHI)and air pollution.Numerous studies have confirmed that there is a coupling relationship between UHI and air pollution,and the superposition of UHI and air pollution makes residents more vulnerable to threats and bear greater risks and pressures.Thus,it is significant to study the spatial coupling relationship between the two phenomena,discuss the influencing factors from the perspective of urban planning adapting to climate change,and seek comprehensive mitigation measures for ameliorating urban ecological environment and improving human settlement environment quality.In this study,Land surface temperature(LST)was retrieved from Landsat8 images,and Aerosol optical depth(AOD)was extracted from MODIS.The spatial coupling relationship of UHI and AOD distribution was studied by using correlation analysis,regional statistics,spatial superposition and coupling model.Then,to seek the major factors that influence their spatial coupling,from the perspective of urban planning to cope with climate change,we selected the indexes about land cover type and spatial structure and conducted correlation analysis and stepwise regression analysis at different scales.The main results were as followed:(1)There were obvious seasonal differences in the intensity and spatial distribution of heat island in Wuhan.The heat island intensity varied with season in the following overall:summer>spring>autumn>winter.In spring and summer,the hot island area was distributed with a large scale,and in autumn and winter it was distributed with a flake or dot shape.In general,the UHI in the study area showed the characteristics of multi-center strong heat source.The urbanized area,industrial zone,main road showed the strong heat island effect,while the natural surface such as water bodies and vegetation area displayed the cold island effect.(2)The AOD values of different seasons in Wuhan were significantly different.The AOD values varied with season in the following overall:winter>autumn>spring>summer.The spatial distribution of AOD was also significantly different.The high AOD values were mainly distributed in the large water bodies,such as the Yangtze River,Zhangdu Lake and Wu Lake,which were in the urban center and sub-urban areas.In spring,AOD values in urban center were lower than those in suburban areas.While in other seasons,the AOD values in urban center were higher than those in the distant urban areas,showing the characteristics of pollution island.(3)The UHI and AOD distribution were coupled in space in different seasons.There was a significant negative correlation between LST and AOD in spring,while significant weak positive correlation between LST and AOD in summer,autumn and winter.The correlation coefficients of spring,summer,autumn and winter were-0.507,0.028,0.159 and0.164,respectively.There was a negative linear relationship between UHI intensity and AOD distribution level in spring and a positive nonlinear relationship in summer,autumn and winter,with the fitting equation R~2was 0.7332,0.6077,0.3455 and 0.7059,respectively.The coupling relationship between UHI and AOD distribution in water bodies was weak,but it was strong in vegetation covered area and urbanized area.(4)We analyzed the effects of land cover characteristics on the spatial coupling between UHI and AOD distribution at different scales.Results show that LST,AOD and its coupling relationship were related to land cover characteristics.The major factors on different scales were different,and the contribution of same factors at different scales was also different.On the 3km*3km grid cell in the urban area and 2km*2km grid cell in the metropolitan area,LST,AOD and its coupling index were promoted by the impervious surface-related variables such as such as NDISI,Percentage of Built-up area and were inhibited by vegetation-related indexes such as NDVI.At the scale of 1km*1km in main urban area,the major factors were different.The contribution of land type composition to the spatial distribution and coupling of LST and AOD was greater than that of land cover structure characteristics.