Effects of Urbanization and Climate Change on Heat Stress Under Relatively Dry and Wet Warm Conditions in a Semiarid Urban Environment

In this work, we investigate urbanization and climate change impacts on summertime heat stress (HS) for Arizona's two largest urban agglomerations, the Phoenix and Tucson metropolitan areas, under relatively dry and moist warm conditions. We perform eight numerical summertime experiments with the Weather Research and Forecasting (WRF)-urban modeling system at high-spatial resolution. We dynamically downscale two contemporary three-month-long summers, one relatively dry and one relatively moist, based on their respective seasonal-mean specific humidity across AZ. Urbanization impacts on HS are assessed by performing two identical numerical experiments for each contemporary summer using different land use-land cover representation: one numerical experiment is performed with the current urban landscape, and one simulation replacing the urban cover with the most representative MODIS-based vegetation type in the region (i.e., open shrub-lands). Climate change impacts on HS are evaluated by performing four additional three-month-long future summer WRF simulations, two via dynamical downscaling of relatively dry conditions (one summer under the RCP8.5 and one summer under the RCP4.5 pathway of greenhouse gas emissions scenario) and two of relatively moist conditions (one summer for each RCP pathway). The selection of future summers is based on their respective seasonal-mean specific humidity across AZ from an end-of-century analysis of 2086-2100. We characterize impacts on HS by examining changes on near-surface air temperature, Heat Index (HI) used by the NWS, and Universal Thermal Climate Index (UTCI) across both megapolitan areas under both dry and moist warm conditions. Our results demonstrate that climate change impacts on HS are not well captured by examining just the projected changes on air temperature and are dependent on the bioclimate index considered. Additionally, we evaluate by means of a human heat balance approach the number of hours per day that an acclimatized (and non-acclimatized) person would experience non-compensable heat stress and compare these results (with the number of hours per day) that we obtain when the HI and UTCI surpass commonly used thresholds considered "dangerous" and of "extreme heat stress", respectively. The HI and UTCI overestimate the number of hours per day that an acclimatized person would experience non-compensable HS under both dry and moist conditions.