The Meteorology of Urban Heat Islands

While urban heat islands have mostly been studied as climatological phenomena, the temporal variability of their signal hints at significant meso to synoptic time-scale dynamics. In addition, health and other impacts of urban overheating tend to be concentrated during regional heatwaves, and mitigation of these impacts thus also needs to focus on their extremes rather than climatic averages. In this talk, we analyze data from Phoenix (collected by NOAA’s Automated Surface Observing Systems - ASOS), focusing on the UHI signal at hourly and daily scales. Analysis of the probability density function of the UHI signal shows distinct dynamics at the hourly versus daily scales, and during daytime versus nighttime. In all cases however, the upper tails are several degrees higher than the mean and the variance is significant, underlining the prominent role of meteorological variability. Attribution analysis of this variability using gradient boosting approaches helps identify the leading drivers (visibility, cloudiness, wind speed and direction, and hour of the day), but the predictive skill of the model is limited, hinting at missing drivers that may be more elusive to quantify or analyze (e.g. anthropogenic heat releases that also vary at hourly scales, inertia/memory effects, among others). We end the talk with an overview of the need to better represent the effect of these fast drivers in numerical simulations of urban meteorology.