Study On The Influencing Factors Of Urban Thermal Environment Based On Visual Index

Cities continue to promote urbanization by expanding built-up area and increasing building height,accompanied by large anthropogenic heat emissions,a process that poses a challenge to the stability of the urban thermal environment.Previous studies have explained the role of urban buildings and greenery(both two-dimensional and three-dimensional)on the urban thermal environment,but their effects are variable across time and seasons.Therefore,this study aims to investigate the combined effects and seasonal differences of urban street constituents on the urban thermal environment.In this thesis,we investigate the effects of urban composition on urban thermal environment parameters(land surface temperature LST,air temperature T_a,relative humidity RH,and temperature humidity index THI)through visual indices(including sky view index SVI,construction view index CVI,road and pavement view index R&PVI,green view index GVI,human view index HVI,and vehicle view index VVI)proposed on the basis of green view index(GVI).Taking Xuzhou City as the research object,we conducted field research on visual indices and urban thermal environment in each season from 2021 to 2022,and obtained visual indices and urban thermal environment distribution characteristics in spring,summer,autumn and winter.Through mobile observation,the image video and urban thermal environment parameters along the sampling route are obtained.Python-Open CV and Seg Net semantic segmentation are used to process and identify the corresponding images to obtain six visual index data.The relationship between the visual indice and the urban thermal environment was explored through correlation analysis.The explanatory power of each visual index on the urban thermal environment was explored through factor detection analysis in the geodetector analysis,and the explanatory power of the interaction between indicators on the changes of the urban thermal environment was explored through two-factor interaction detection.The independent and joint contribution of the visual indices to the changes of the urban thermal environment was explored through hierarchical analysis.Based on the statistical analysis,the possible causes were discussed in the context of the actual situation and previous studies,and the influence of visual indices on the urban thermal environment in different seasons was summarized to explore the differences and changes of the dominant factors,and the main conclusions were as follows:(1)In terms of spatial distribution characteristics:within the main urban area(urban expressways and urban roads)has higher CVI and R&PVI,while LST and T_aare higher,and there are obvious characteristics of urban heat island effect;inside the scenic area,on the contrary,there is a distribution of higher GVI,lower T_a and higher RH,and the"cold island effect",the characteristics of"cold island effect"are obvious and the thermal comfort is also higher.In addition,the values of SVI,CVI,R&PVI and GVI have obvious seasonal characteristics,while HVI and VVI have the lowest proportion among the six indices,mainly because the anthropogenic heat sources along the sampling route are more dispersed;the values of urban thermal environment parameters LST,T_a,RH and THI are also different in each season,but for THI,except for summer,it is comfortable state.(2)Differential effects of different time periods:The correlation and contribution of morning time are higher than that of afternoon in all seasons,i.e.,the visual indices in morning time are more capable of explaining the spatial distribution of urban thermal environment parameters.(3)The main influencing factors of the urban thermal environment:influenced by seasonal differences and inconsistency of human activities,the influencing factors of the urban thermal environment show seasonal characteristics.In spring,the main influencing factors for LST and RH are CVI and GVI,and for T_a and THI are R&PVI and HVI,respectively.In summer,the main influencing factors for LST,T_a,RH and THI are GVI and HVI.In autumn,the main influencing factors are CVI and GVI,and in winter,the main influencing factors are CVI,GVI,HVI and VVI.Among them,the individual influence of SVI is weak,but the interaction with other factors in the interaction detection analysis has significantly increased the explanatory power of the urban thermal environment.(4)Combined effects of different seasonal drivers:The geoprobe analysis shows that the interactions among the visual index factors have enhanced explanatory power for the seasonal urban thermal environment,and the interactions are all bifactor-enhanced or non-linearly enhanced,indicating that the spatial differences in the urban thermal environment are caused by the combined effects among building morphology,ground cover(including greenery,hardened ground surface,etc.),and human activities.(5)Seasonal influence pattern of visual index on urban thermal environment:Seasonal influence pattern of visual index on outdoor thermal environment:SVI and LST are mainly negatively correlated in spring,positively correlated in summer,but not significantly correlated in autumn and winter;to T_a is mainly negatively correlated in spring and winter,and shows a shift from negative correlation to positive off in summer and autumn;to RH in spring,summer and winter are positive correlation in spring,summer and winter,but not in autumn;with THI,negative correlation was predominant in spring and winter,but positive correlation was predominant in summer and autumn.CVI was positively correlated with LST in spring and winter,and the modulating effect was most obvious in winter;it was positively correlated with T_a in all seasons,and was also stably negatively correlated with RH,and was mainly positively correlated with THI.R&PVI and LST were negatively correlated in the morning and positively correlated in the afternoon in both spring and summer,and were mainly negatively correlated in autumn,but not in winter;and T_a experienced a shift from negative correlation in the morning to positive correlation in the afternoon in spring and summer,and were not correlated in autumn and winter.GVI was negatively correlated with LST in spring and winter,and no significant correlation in summer and autumn;it was a stable cooling effect for T_a,and a stable humidifying effect for RH,significantly improving human thermal comfort.HVI and VVI were mainly positively correlated with LST,T_a and THI in all seasons,and negatively correlated with RH,and their warming effects were most obvious in winter.