Abstract
Research was insufficient on the surface urban heat islands (SUHIs) in China during the morning and before midnight. Therefore, we studied the spatio-temporal variations and ten determinants of SUHI intensities (SUHIIs) at ∼10:30 am and ∼10:30 pm of 932 urban region agglomerations in the five ecological regions in China and further simulate these SUHIIs using the machine learning algorithm. The main findings were as follows. (1) The daytime SUHIIs in all regions were largest in the summer, and when the humid regions had higher SUHIIs than other regions. The maximum SUHIIs were usually significantly positively correlated with the minimum ones. (2) The SUHIIs were significantly partially correlated with the enhanced vegetation index, nighttime light intensity, PM2.5, albedo, population density, precipitation, urban area size and landscape shape index in many or a few cases. (3) The root-mean-square errors of simulated daytime and nighttime SUHIIs were mostly less than 1.22 and 2.00 °C, respectively.
Data availability
The data in this study are available from the corresponding author upon request.
Disclosure statement
The authors have no relevant financial or non-financial interests to disclose.