Transfer learning reveals large discrepancies between air and land surface temperatures in cities

Understanding of urban weather and climate is severely limited by data poverty resulting from a dearth of true urban weather stations ^1,2 . As a result, land surface temperature ( T _s ), obtained from remote sensing platforms, has been widely used as a stand-in for near-surface air temperature ( T _a ) ^3–7 despite their fundamental differences ⁸ , especially in urban areas. This has led to erroneous characterization of urban heat stress and urban climate impacts ⁹ . Here we develop a novel urban transfer learning framework (U-TL) to address this critical gap and to provide urban high-resolution air temperature (U-HAT) data at large scales across the contiguous United States (CONUS). U-TL demonstrates high accuracy and strong robustness in predicting urban T _a , even with limited training data. The resulting U-HAT is the first high-resolution urban T _a dataset capable of accurately reproducing observed and well-established urban climatology. U-HAT reveals substantial T _s –T _a discrepancies and therefore cautions the use of T _s to characterize urban heat. We show that satellite-measured T _s substantially overestimates both urban heat stress magnitude and intra-city disparities, which have very consequential implications for urban heat exposure, vulnerability, and adaptation policy making.