Linking spatiotemporal variations in urban land surface temperature to land use and land Cover: A case study in Hangzhou City, China

Urbanization-induced Land Use and Land Cover (LULC) changes are considered the direct cause of Urban Heat Island (UHI), whereas the impact of LULC on Land Surface Temperature (LST) exhibits spatiotemporal variability. Investigating the key variables controlling LST in urban areas with significant landscape heterogeneity remains challenging. This study focuses on the relationship between UHI and LULC in Hangzhou, China. Four Landsat-8 images, spanning consecutive seasons from 2015 to 2016, were used to retrieve LSTs, Normalized Difference Vegetation Index (NDVI), and Normalized Difference Moisture Index (NDMI). We employed correlation analysis, redundancy analysis, and both the ordinary least squares (OLS) linear regression and Geographically Weighted Regression (GWR) to link spatiotemporal variations in LST to LULC. The main findings are as follows: (1) the seasonal influence of LULC on LST exhibits a trend of stronger effects during warmer seasons and weaker effects during colder seasons; (2) compared to the proportion of vegetation landscapes, the proportions of water bodies and built-up areas have a significantly greater impact on LST. Water bodies exert the strongest influence on LST and should be prioritized in UHI mitigation planning; (3) NDMI consistently reflects surface moisture conditions across various land cover types more effectively than NDVI represents vegetation coverage, resulting in a stronger and more stable negative correlation with LST across all seasons; and (4) the GWR approach, which accounts for the spatial non-stationarity of the LULC-LST relationship, reveals that NDMI models outperform NDVI and landscape proportion models, offering better explanatory power for LST at local scales.