Publications

Ge, S; Zhan, WF; Wang, SS; Du, HL; Liu, ZH; Wang, CG; Wang, CL; Jiang, SD; Dong, P (2024). Spatiotemporal heterogeneity in global urban surface warming. REMOTE SENSING OF ENVIRONMENT, 305, 114081.

Abstract
The rapid urban warming in recent decades has posed significant risks to the health and well-being of urban residents. Previous studies have predominantly examined urban surface warming from an annual-mean and whole-city perspective. Spatiotemporal heterogeneity in surface warming trends within cities and throughout different periods within a yearly cycle remains largely unclear across global cities. Here we investigate the spatiotemporal heterogeneity in both daytime and nighttime urban surface warming trends (2003-2020) over 515 cities worldwide, mainly based on satellite-derived land surface temperature (LST) data. Our results reveal a distinct pattern in the interannual daytime and nighttime LST trends (delta LST), characterized by an initial rise and subsequent decline from city centers to the outskirts. Daytime delta LST is more prominent in urban expansion zones near the edge of the city rather than at the outermost city edges, with an average trend reaching 0.072 +/- 0.062 C-degrees/yr. These regions exhibit a robust negative correlation between daytime delta LST and declining trends in Enhanced Vegetation Index (r = -0.30, p < 0.05), suggesting accelerated warming associated with greenery reduction during urbanization. Nighttime warming also peaks in urban expansion zones, with an average trend of 0.058 +/- 0.031 C-degrees/yr. Nonetheless, the zones where nighttime warming peaks are closer to city centers compared to those during daytime. Nighttime delta LST is strongly correlated with decreasing albedo. Regarding temporal heterogeneity, our analysis for cities in the Northern Hemisphere indicates that delta LST peaks in February (day: 0.113 +/- 0.123 C-degrees/yr; night: 0.099 +/- 0.097 C-degrees/yr) and reaches its lowest point in November (day: 0.010 +/- 0.078 C-degrees/yr; night: 0.004 +/- 0.079 C-degrees/yr). Statistical analysis reveals that monthly delta LST is more regulated by climatic factors than surface properties. In particular, monthly delta LST shows a positive correlation with background surface air temperature trends (avg. r = 0.56, p < 0.05). For cities in the Southern Hemisphere, the fastest daytime delta LST occurs in January at 0.075 +/- 0.095 C-degrees/yr, while the quickest nighttime delta LST occurs in May at 0.057 +/- 0.048 C-degrees/yr. Daytime delta LST is equally regulated by climatic factors and surface properties. However, nighttime delta LST is influenced more by climatic factors. We believe this study advances our understanding of the spatiotemporal heterogeneity of global urban surface warming trends and provides valuable insights for identifying intra-city heat risks.

DOI:
10.1016/j.rse.2024.114081

ISSN:
1879-0704