The Influence of 2D/3D Urban Spatial Form Indicators on Surface Urban Heat Island Based on Spatial Regression Models: A Case Study of Hangzhou, China
Haotian Chen,Sheng Zheng
DOI: https://doi.org/10.1007/978-981-99-3626-7_65
2023-01-01
Abstract:Rapid urbanization in China has caused the serious urban heat island (UHI) effect, which is endangering the environment where people live. Many studies have demonstrated the significant contributions of two-dimensional (2D) and three-dimensional (3D) urban spatial form indicators to UHI, but further research is expected to conduct given that UHI has spatial non-stationarity and spatial spillover effects of explanatory variables. Spatial regression models are particularly appropriate to solve the spatial non-stationarity problem. This essay seeks to analyze the impact of urban spatial form on UHI taking the spatial spillover effects into account. In this paper, Landsat 8 OLR/TIRS satellite image was used to invert surface urban heat island intensity (SUHII) in the central city of Hangzhou, China. The Moran's I index for SUHII was 0.457, indicating a substantial positive spatial autocorrelation. After the establishment of 2D/3D urban spatial form indicators at the block scale level, the influence of 2D/3D urban spatial form indicators on SUHII w`as examined using the spatial lag model (SLM), spatial error model (SEM), spatial Dubin model (SDM), and spatial Dubin error model (SDEM), respectively. Results suggest that SDM and SDEM fit better by considering spatial interaction of urban spatial form indicators. The coefficients for the ratio of impervious surface area (RISA), ABD (average building density), and sky view factors (SVF) were significantly positive, while the effects of the landscape shape index (LSI), average building height (ABH), average forest height (AFH), and normalized difference vegetation index (NDVI) were significantly negative on the SUHII. Additionally, there were spatial spillover effects of LSI, SVF, AFH, NDVI, and SUHII affecting SUHII. The findings imply that the urban form indicators and SUHII will affect the SUHII of adjacent blocks, so it is important to make suitable urban 2D/3D shapes to enhance the urban thermal environment. These findings would be helpful for urban planners to mitigate future UHI effects.