Spatial expansion paths of urban heat islands in Chinese cities: Analysis from a dynamic topological perspective for the improvement of climate resilience

• The urban heat island expansion index was developed to identify the new UHI patches. • The new UHI patches are reclassified as infilling, edge-expanding and leapfrogging. • The new UHI patches have lower UHII but higher contributions than pre-existing patches. • The new UHI patches are predominantly edge-expanding with the maximum UHII. • Croplands, rural residential areas, and forests were the main sources for new UHI patches. Urban ecological resilience enhancement is important for the mitigation of climate change risks; however, urban heat islands (UHIs) have negative impacts on urban resilience. A UHI expansion index (UHIEI) was developed to identify new UHI patches produced by including infilling, edge expansion, and leapfrogging. In this study, we used a simplified urban extent (SUE) algorithm to estimate new UHI patches and their associated UHI intensity (UHII) values in 371 cities in China during 2005–2020. Then, we analyzed the sources and sinks of the UHI patches according to land use data. The study sought to comprehensively determine the spatial expansion path of the new UHI patches at the national scale. The results showed mean UHII values of 2.11 °C ± 0.63 °C and 1.06 °C ± 0.54 °C during the day and night, respectively, for the 371 cities in the summer of 2020; these were slightly higher than the corresponding values in 2005. The UHII values of the new UHI patches were 0.57 °C and 0.29 °C lower during the day and night, respectively, compared with pre-existing UHI patches in summer. New UHI patches were predominantly formed through edge expansion, with maximum UHII values of 1.74 °C ± 0.80 °C and 0.88 °C ± 0.42 °C during the day and night, respectively; these are lower than the values of pre-existing UHI patches, but they represent greater contributions to the land surface temperature (LST) of the city. Croplands, rural residential areas, and forests were the main sources for new UHI patches. The results of this study will allow better identification and comparison of the temporal and spatial characteristics of pre-existing and new UHI patches; they will also facilitate the design of targeted measures to mitigate their ecological impacts according to expansion type, thereby improving the cities’ ecological resilience characteristics.