Spatiotemporal evolution characteristics and driving factors of heat island effect based on territorial perspective: A case study of Beibu Gulf urban agglomeration, China

The heat island effect has far-reaching impacts on human settlements and natural ecosystems. Current research predominantly focuses on urban areas and overlooks the broader natural ecosystems beyond urban boundaries. Therefore, there's an increasing urgency to reassess and study the heat island effect from a regional spatial perspective. This study introduces the concept of Land Heat Island (LHI) and quantifies it through Land Heat Island Intensity (LHII). Utilizing multivariate data and statistical analyses, this study investigates the spatiotemporal evolution of LHII in the Beibu Gulf urban agglomeration from 2012 to 2021. Redundancy analysis (RDA) and spatial bivariate correlation analysis were used to understand the relationships and evolution of factors influencing LHII. The results showed: (1) From 2012 to 2021, the heat island effect within the Beibu Gulf urban agglomeration gradually intensified. LHII was higher in the coastal areas compared to the inland regions, more developed economic areas exhibited higher LHII than less developed ones, and areas characterized by concentrated human activities had higher LHII than ecological agglomeration areas. (2) The primary zones of LHII growth in the Beibu Gulf urban agglomeration were slightly hot island areas and weakly hot island areas. The northern part of the area where forestland was clustered had a greater influence on the cold island pattern than on the heat island pattern; the lower latitude area had a greater influence on the heat island pattern than on the cold island pattern; and the eastern part of the area where the economy was well developed had an increasing influence on the heat island pattern year by year. (3) The enhancement of LHII in the Beibu Gulf urban agglomeration is primarily driven by urban construction and socio-economic factors. Specifically, Carbon Emissions and the Nighttime Light Index (NLI), which representing the strength of urban construction, contribute 11.4% and 8.3% respectively. Additionally, the Population Size (PS), representing the strength of socio-economic factors, contributes 12%. (4) The weakening of LHII was primarily affected by natural ecological factors, with the top two contributors being the average Digital Elevation Model (DEM) of cultivated land and average Normalized Difference Vegetation Index (NDVI), accounting for 32.9% and 14% respectively. This study enhances our understanding of the spatiotemporal evolution of the heat island effect and aims to mitigate it in the Bay Area urban agglomeration.