Divergent Heat Assessments Across Thermal Stress and Sensation Metrics

Abstract

A large and growing body of literature has indicated potential inconsistencies in heat assessments due to the use of different heat stress indices (HSI). This divergence hinders the emergence of scientific consensus and impairs the development of public policy on a range of topics, including climate change impacts, public health guidance, and heat adaptation strategies. Beyond discrepancies among physically based HSI themselves, divergence may also arise between physical thermal stress and psychological thermal sensation; such deviation can delay active behavior adaptation measures but has rarely been examined. To explain these metric divergences, we first connect the different assessments of HSI to their varying sensitivities to meteorological drivers and introduce a clustering framework to guide their applications. We then identify spatiotemporal hotspots where sensation-based estimates diverge from HSI-based assessments, highlighting cases where reliance on sensation alone may increase risk. Particularly, nighttime humid heat in low-latitude coastal Asian cities is perceived as less risky than indicated by HSI due to a muted perception of high humidity impact and the absence of solar radiation that would strongly signal heat discomfort. These findings help elucidate inconsistent conclusions in the existing literature stemming from the use of different thermal metrics and localize spatiotemporal hotspots for targeted heat health warning systems.