Evaluation of urban heat island effects based on fine-resolution mapping of subpixel impervious surface dynamics over four cities in China

• We propose UISP for subpixel impervious surface extraction using Landsat imagery • UISP reveals urban sprawl and enhances SUHI analysis beyond pixel-scale limits • High ISP values are strongly associated with intensified urban heat island effects • UISP supports refined urban heat management and resilient city development policies In the context of global urbanization, urban expansion significantly affects urban thermal environment. As a key indicator of urbanization, impervious surfaces are widely used to monitor urban growth. However, most existing studies examining thermal environments through impervious surfaces are limited to pixel-scale. Given the complexity and heterogeneity of urban land cover, pixel-scale analyses often fail to accurately capture the thermal effects of impervious surfaces. To address this issue, we propose a universal impervious surface percentage (UISP) model based on Landsat imagery for subpixel-scale impervious surfaces monitoring. We evaluated the relationship between impervious surface percentage (ISP) and surface urban heat islands (SUHI) in four representative Chinese cities—Beijing, Kunming, Shenzhen, and Wuhan—from 2018 to 2023. Results indicate that UISP significantly improves the accuracy of impervious surface estimation over heterogeneous urban landscapes. The deviations of estimated ISP are as much as 0.2 lower than pixel-scale results. In SUHI areas, ISP values exceeded 0.4 to 0.6 higher than in non-SUHI regions. The ISP difference between SUHI levels 1 and 2 was 0.12, and between levels 5 and 6, it was 0.08, suggesting ISP saturation in high SUHI zones. This study introduced UISP for subpixel extraction of impervious surfaces and thermal environment analysis, offering a novel approach for precise quantification of urbanization and insights into thermal environment research.