Driving mechanism of thermal environment based on Bayesian structural equation model

Abstract

The urban thermal environment has an important impact on human health and sustainable development. Most of the current mainstream studies on the influence of thermal environment focus on the relationship analysis with a single factor, but there are fewer studies on multiple factors and their interactions, and there is a lack of analysis of human activities as an important factor. In this paper, the relationship between summer surface temperature and natural environment, surface cover, landscape structure, and human activities in Beijing-Tianjin-Hebei region is investigated in depth by using Bayesian structural equation model. The results show that: impervious surface (0.66) and air temperature (0.78) have significant positive effects on the thermal environment, while vegetation (-0.37) and transpiration (-0.27) have significant negative effects on the thermal environment; human activities indirectly affect the surface thermal environment through their effects on the natural environment and landscape structure, so the construction of the driving mechanism model of the surface thermal environment needs to consider the role of human activities The model results are optimal at 6k scales, with Rhat less than 1.01 for all parameters, Bulk_ESS and Tail_ESS values greater than 1500, and no effect value over 1. The LOOIC value is 67431.1. The remote sensing image data has the advantages of sufficient data volume and significant spatial and temporal effects, so it is feasible and effective to apply it to BSEM analysis.