Mapping individual urban trees in arid cities: linking canopy coverage to land surface temperature variability using remote sensing

The spatial distribution and canopy traits of urban trees regulate their ability to support ecosystem functions and societal needs in urban environments. However, the distribution of individual trees in arid urban settings and their association with land surface temperatures remain unclear. By combining Google satellite imagery and Gaofen-2 imagery, we present a “position-constraint-segmentation” method for identifying the tree crown cover of individual trees in arid urban environments. We implemented the approach in five cities in Northwest China, including one city with a continental plateau semiarid climate and four arid cities. We investigated the relationship between tree crown cover and land surface temperatures, derived from Landsat 8 imagery, across five cities. The model demonstrated a robust accuracy with an overall correlation coefficient Formula: see text of 0.74 (p < 0.001) between manually labeled and predicted tree counts across randomly selected image blocks. We identified a total of ~9.1 × 105 individual trees in five representative cities, including ~5.6 × 105 in Xining city. Tree crown size exhibits a unimodal distribution with a mean value of 63.7 m2, and 12.69% of trees have crown cover greater than 100 m2 in Xining city. Land surface temperatures exhibit significant variations across different urban land use types, with the lowest temperatures observed in parks and greenspaces, and the highest in commercial service areas. Tree clusters and strategic tree retention could effectively mitigate urban heat accumulation. This study presents novel individual-tree detection methods for arid cities using high-resolution remote sensing imagery, delivering fine-grained data for urban resource management beyond coarse-scale approaches. Future efforts should promote the construction of intelligent greening system adapted to climate change, and contribute to sustainable development in arid cities.