Assessing the thermal contributions of urban land cover types

Understanding the thermal contribution of urban land cover is crucial for alleviating urban heat islands (UHIs). Extensive work has assessed this contribution by estimating the responses of heat-related variables, such as land surface temperature (LST), to landscape patterns in terms of composition and configuration. However, ignoring the endogenous collinearity in landscape composition may lead to biased estimations. In this study, an elastic net regularized regression was used to disentangle the thermal contributions (i.e, the local cooling/warming effects) of six urban land cover types (i.e., water body, urban tree, grassland, bare land, impervious surface, and building) in the Beijing metropolitan region of China. In addition, the benefits of cooling/warming gains from the spatial aggregation of vegetation/buildings were quantified. The results indicate that for a 10% increase in coverage within an area of 1440 × 1440 m2, buildings appear to have a strong warming effect and raise local LST by ~1.26 ℃, which is much higher than the warming caused by impervious surfaces (~0.23 ℃). In contrast, water bodies, grasslands, and urban trees have different cooling effects that reduce local LST by ~0.72 ℃, ~0.60 ℃ and ~0.57 ℃, respectively. Landscape configuration interactively affects local LST based on the composition. The aggregation cooling (~-1.2 ℃ at maximum) of vegetation only takes effect when the local vegetation coverage is less than 40%. The aggregation warming (~1.3 ℃ at maximum) of buildings occurs when the local building coverage is more than 15%. These findings provide new insights into the thermal contribution assessment of urban land cover that helps create urban heat island (UHI) mitigation strategies and plan future landscapes.