<b>River </b><b>Cooling </b><b>Distances</b><b> of the Xiang River in Hengyang: Spatiotemporal Evolution and Associated Factors Based on LCZ</b>

We retrieved Land Surface Temperature (LST) from Landsat imagery and mapped Local Climate Zones (LCZs) using the Random Forest algorithm. By integrating multi-ring buffer analysis with the Geodetector model, we quantified the river cooling distance across different LCZ types along the urban reach of the Xiang River in Hengyang City.

Our results indicated that: (1) surface thermal conditions exhibited marked cross-LCZ thermal contrasts among the three summer acquisitions, with the most prominent warming anomalies found in LCZ 3, 7, 8, and 10; (2) the river cooling distances of built-up LCZs were 820 m (2011), 970 m (2018), and 780 m (2025), all falling within the 750–1000 m range; (3) the primary spatially associated factors shifted from built-up intensity (NDBI) to nighttime light (NTL) and population density (POP).