Identification and quantification of the effects of source and sink areas on land surface temperature in Shanghai

Abstract:

The effect of the urban heat island (UHI) is caused by landscape change in urban environment. Understanding the relationship between urban heat island and landscape pattern is important for UHI mitigation. We studied the source and sink landscapes in Shanghai to investigate the heterogeneous effects of landscape patterns on land surface temperature (LST) using semiparametric geographically weighted regression (SGWR) models. The results indicated that landscape composition had stronger effects on thermal environment than configuration. Specific findings include: (1) The general pattern of LST in Shanghai is higher in the north and lower in the south; (2) The high LST zones were associated with including air and other ports around the city; the sub-high LST zones were mainly distributed in buildings areas while the sub-low and low LST zones were mainly in the west with rich vegetation; (3) A contribution index (CI) was defined and used to represent the contribution of land cover to the thermal environment; Results suggested that the DI values of construction land (1.854) and bare land (1.104) were significantly larger than 1, implying aggravation of the UHI effect as heat sources. In contrast, cultivated land (0.084), forest land (0.073), grassland (0.615), and water bodies (0.098), each had DI value less than 1, were identified as heat sinks in 2017. The CI provides a convenient way to measure the contribution of urban landscapes to the overall UHI which can be applied to other cities to quantify the effects of landscape composition and configuration on UHI.