Driving mechanism of urban heat island spread in the central-southern Liaoning urban agglomerations, China (2013–2020)

Urban Heat Island (UHI) spread has transitioned from a localized phenomenon to a region-wide challenge. However, previous research focusing on individual cities limits understanding of how UHI propagates across city clusters, where socio-economic and natural factors interact across administrative boundaries. This study examines cross-boundary drivers of UHI spread in the Central-southern Liaoning Urban Agglomerations (CSLUA) from 2013 to 2020 using spatial econometric models. The results reveal that: (1) The reduction in non-UHI areas in CSLUA was accompanied by substantial expansion of moderate and strong heat island zones. Urban heat does not respect administrative boundaries, and urban cores function as "thermal hubs," exporting heat to adjacent areas through spillover effects. (2) Vegetation demonstrates paradoxical effects: NDVI has a negative direct effect on UHI locally, but a positive indirect effect on UHI in adjacent areas. This positive indirect effect does notindicate high NDVI counties aggravate their neighbors; it reflects that high NDVI counties receive heat pressure from neighboring urbanized counties with low NDVI, high temperatures, dense impervious surfaces, and intensive development. (3) Population density has a positive indirect effect on UHI, while GDP exhibits a negative indirect effect, suggesting mitigation via industrial restructuring and ecological investment. These findings provide new insights into the drivers of cross-city UHI spread and call for a shift from city-centric greening to agglomeration-scale land use governance. Prioritizing forest-water corridor networks, ecological infrastructure, and coordinated population relocation is essential to decouple urban growth from thermal intensification and build climate-resilient urban systems. • UHI in CSLUA shows a significant increasing trend during 2013–2020. • NDVI cools locally but shows positive spillovers via thermal encroachment. • Population density raises LST indirectly, while GDP reduces it via spillovers.