Login / Signup

The development of an adaptive heat stress compensability classification applied to the United States.

Gisel Guzman-EchavarriaAriane MiddelDaniel J VecellioJennifer Vanos
Published in: International journal of biometeorology (2024)
Traditional climate classification and weather typing systems are not designed to understand and prevent heat illness, or to design effective cooling strategies during extreme heat. Thus, we developed the Heat Stress Compensability Classification (HSCC) combining open-source historical weather data (2005-2020) with biophysical modeling of a standard human, in the sun or shade, during peak city-specific hot hours on the top 10th percentile hottest days in 96 U.S. cities. Four categories of uncompensable heat stress (UHS)--which can result in rising core temperature--were established based on the relative constraints of dry and evaporative heat exchanges for achieving heat balance in proportion to constant metabolic heat production (112Wm -2 ). Results show that 88.7% of these peak-hot hours meet the UHS criterion, and 41% present a dry heat gain of 70 to 150Wm -2 while allowing a maximum evaporative loss between 90 and 140Wm -2 . Evaporative heat loss constraints dominate the eastern U.S. Dry heat gain was widespread, yet particularly high in the south and southwest. Full shade reduces UHS frequency to 7.6%, highlighting the importance of quality shade access and accounting for radiative load in heat stress assessments. Although there are five distinct categories (one compensable and four UHS), the HSCC is dynamic and customizable, providing actionable information on thermal variations within a given category. These variations depict the reason for UHS (e.g., limited evaporative cooling) and, thus, how to concentrate cooling efforts, particularly at the limits of physiological adaptability. Findings facilitate developing targeted criteria for heat stress reduction with potential global applications.
Keyphrases
  • heat stress
  • heat shock
  • machine learning
  • deep learning
  • endothelial cells
  • electronic health record
  • quality improvement