Quantifying the Microscale Details of Wet Bulb Globe Temperature Across Urban Heat Islands

Heat is the leading cause of weather-related fatalities worldwide and is intensifying due to both anthropogenic climate change and urbanization. Wet Bulb Globe Temperature (WBGT) is a heat stress index that is becoming the most widely used assessment of heat stress, as it accounts for the combined effects of air temperature, humidity, wind, and solar radiation on the human body during physical activity. However, WBGT weather station observations are sparse, and forecasting is often limited to coarse spatial scales (2.5 km × 2.5 km). WBGT varies widely across much smaller spatial scales, influenced by the land surfaces that modify localized values of air temperature, shading, ventilation, and humidity. In this dissertation, mobile transect observations and machine learning are used to quantify heat stress across a variety of land surfaces. Microscale variations in WBGT were quantified using data from mobile vehicle, cycling, and walking observations and modeled across Carrboro, Chapel Hill, and South Durham, North Carolina, with machine learning methods. Using these data, the fine-scale influence of localized land surface characteristics on WBGT across a range of heat-season days and weather conditions is quantified. This work revealed that the spatial distribution and intensity of WBGT are strongly influenced by land-surface characteristics, land-surface composition, and concurrent weather conditions. Differences among land-cover classes (e.g., forested vs. developed areas) frequently produced simultaneous, varying levels of heat risk within the same observation period. Wider disparities in heat stress between forested and developed areas were observed on hot, sunny, calm, and dry days, while breezes and cloud cover played a disproportionately large role in cooling highly developed areas. Notably, the greatest heat-stress contrasts did not emerge within expansive areas of uniform impervious or forest cover but rather in small- to moderately sized zones of land-surface transition. Finally, our findings underscore that WBGT is a complex metric with pronounced microscale variability. WBGT should be regarded as a spatially heterogeneous metric whose magnitude and distribution can shift substantially in response to relatively small changes in weather conditions or land-cover patterns.