Cloud probability distribution of typical urban agglomerations in China based on Sentinel-2 satellite remote sensing

Abstract

Cloud distribution significantly impacts global climate change, ecosystem health, urban environments, and satellite remote sensing observations. However, past research has primarily focused on the meteorological characteristics of clouds with limitations in scale and resolution, leading to an insufficient understanding of large-scale cloud distribution and its relationship with land surface cover and urbanization. This study investigates the cloud distribution characteristics of typical urban agglomerations in different climatic regions of China using high-resolution Sentinel-2 satellite imagery and the Google Earth Engine platform. A cloud probability descriptor was constructed based on three years of high spatiotemporal resolution observations. The results revealed significant differences in cloud distribution among urban agglomerations, challenging the traditional understanding based on climate zoning. The Northeast urban agglomeration in the temperate zone exhibited high cloud coverage (37.54%), while the Chengdu-Chongqing urban agglomeration in the subtropical zone and the Qinghai-Tibet Plateau urban agglomeration in the plateau climate zone had even higher average cloud probabilities (50.72% and 43.27%, respectively). The analysis suggests land surface cover, urbanization, and other surface factors may influence cloud distribution patterns. These findings emphasize the ubiquity of cloud cover and highlight the importance of considering the complex interactions among geographical factors when characterizing cloud cover diversity. This study contributes to providing new insights for enhancing meteorological models and remote sensing observation strategies in cloudy environments across different climate zones.