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

Abstract

Understanding of urban weather and extreme events remains severely limited by data poverty resulting from a dearth of true urban weather stations. 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) despite their fundamental differences, especially in urban areas. Although T_s provides important scientific insights and practical utility for studying the urban thermal environment, this substitution risks introducing large uncertainties and biased characterization of impact-relevant urban heat stress. Here we develop an 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 a 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 spatial variability, which have consequential implications for urban heat exposure, vulnerability, and adaptation policy making.