Quantifying the relationship between land use features and intra-surface urban heat island effect: Study on downtown Shanghai

Hao Zhang,Tian-Tian Li,Jie-Jie Han

DOI: https://doi.org/10.1016/j.apgeog.2020.102305

IF: 4.732

2020-12-01

Applied Geography

Abstract:<p>Fine-resolution products of land use features (LUFs) and the thermally sharpened infrared data are vital to reveal the land use patterns and their influence on the spatial variability of intra-surface urban heat island (SUHI) effect within cities. In this study, 146 land parcels within downtown Shanghai, China, were selected to investigate the relationship between parcel-based LUFs and intra-SUHI effect through the estimated sensible heat flux approach, by combining a thermal sharpening method and statistical analysis. The results show that the parcels with a dominant OpenSpace exhibited the lowest net blackbody sensible heat flux density (NetBBFD) across the study period. The well-planned parcels featured with high-rise and OpenSpace exhibited lower NetBBFD. In contrast, parcels with dominant categories of low-rise, pavement, other LUFs, and their mosaic, exhibited higher NetBBFD. Furthermore, partial least squares regression (PLSR) models reveal a statistically significant relationship between the Box-Cox transformed parcel-based LUFs and net blackbody sensible heat flux (NetBBF). Variations in the predictors explained approximately 65.7–68.7% of the variance in the responses on hot and warm days, whereas variations of the predictors explained approximately 28.7–28.8% of the variance of the responses on cool and chilly days. The PLSR models also quantified the relative importance of univariate effects and the interactions of changing predictors in terms of their contributions to changing responses. Finally, the uncertainties arising from the datasets and methods as well as the implications for sustainable land development and mitigating the UHI effect were discussed.</p>

geography