Remote sensing image-based analysis of the urban heat island effect in Shenzhen, China

Abstract

The urban heat island effect is a typical urban climate phenomenon, and the formation and intensity of these phenomena are closely related to the land cover type. Based on Landsat 8 remote sensing images of Shenzhen, China, this study uses a linear spectral mixture model to extract land coverage information, and then analyzes the effects of urban land coverage on the land surface temperature (LST) and heat budget components. The results show that the LST of Shenzhen decreases gradually from northwest to southeast and that the LST of urban impervious surfaces are significantly higher than those of vegetation-covered areas with a difference of 3–4 K, thus demonstrating a significant urban heat island effect. Impervious surfaces and urban vegetation both affect the LST and heat budget components of Shenzhen. The impervious area contributes more to the surface sensible heat, while urban vegetation contributes more to the surface latent heat. The research results show that a very small area of urban green land (vegetation fraction coverage less than 4–8%) exhibits a strong capacity for heat dissipation. In the practice of urban construction, avoiding concentrated impervious areas and increasing urban green land areas represent effective methods of alleviating the urban heat island effect. This study helps to better understand the optimization of urban landscapes and alleviation of the urban heat island effect.