Attributing the Urban-Rural Contrast of Heat Stress Simulated by a Global Model

The two-resistance mechanism (TRM) attribution method, which was designed to analyze the urban-rural contrast of temperature, is improved to study the urban-rural contrast of heat stress. The improved method can be applied to diagnosing any heat stress index that is a function of temperature and humidity. As an example, in this study we use it to analyze the summertime urban-rural contrast of simplified wet bulb globe temperature (SWBGT) simulated by the Geo-physical Fluid Dynamics Laboratory land model coupled with an urban canopy model. We find that the urban-rural con-trast of SWBGT is primarily caused by the lack of evapotranspiration in urban areas during the daytime and the release of heat storage during the nighttime, with the urban-rural differences in aerodynamic features playing either positive or nega-tive roles depending on the background climate. Compared to the magnitude of the urban-rural contrast of temperature, the magnitude of the urban-rural contrast of SWBGT is damped due to the moisture deficits in urban areas. We further find that the urban-rural contrast of 2-m air temperature/SWBGT is fundamentally different from that of canopy air tem-perature/SWBGT. Turbulent mixing in the surface layer leads to much smaller urban-rural contrasts of 2-m air tempera-ture/SWBGT than their canopy air counterparts. SIGNIFICANCE STATEMENT: Heat leads to serious public health concerns, but urban and rural areas have differ-ent levels of heat stress. Our study explains the magnitude and pattern of the simulated urban-rural contrast in heat stress at the global scale and improves an attribution method to quantify which biophysical processes are mostly respon-sible for the simulated urban-rural contrast in heat stress. We highlight two well-known causes of higher heat stress in cities: the lack of evapotranspiration and the stronger release of heat storage. Meanwhile, we draw attention to the veg-etation types in rural areas, which determine the urban-rural difference in surface roughness and significantly affect the urban-rural difference in heat stress. Last, we find the urban-rural contrasts of 2-m air temperature/SWBGT are largely reduced relative to their canopy air counterparts due to the turbulent mixing effect.