Seasonality of the urban heat island effect: patterns and drivers

Abstract:

We conducted a rigorous analysis of the drivers of seasonality in the urban heat island (UHI) effect. Many studies report annual cycles in UHI intensity and have attributed those patterns to various hypotheses, including seasonal trends in wind and clouds, prevalence of anti-cyclonic conditions, soil moisture, and day length. But to our knowledge, those hypotheses have never been tested, leaving a substantial gap in our basic understanding of the urban climate.

We tested these and other hypotheses using two years of continuous temperature measurements from an array of 150 sensors in and around Madison, Wisconsin USA, an urban area of 407,000 surrounded by lakes and a rural landscape of agriculture, forests, wetlands, and grasslands. This is one of the best replicated urban climate arrays ever deployed, which allowed us to characterize the UHI in rich spatial and temporal detail and rigorously assess the biophysical and synoptic drivers of its seasonal variation.

UHI intensities were typically highest in summer and lowest in winter. Seasonal trends in wind speed and cloud cover generally tracked annual trends in UHI intensity, with the clearer, calmer conditions conducive to stronger UHIs more common in summer. This is consistent with the hypothesis that seasonal trends in wind, clouds, and anti-cyclonic conditions drive UHI seasonality. However, clear, calm summer nights still had higher UHI intensities than clear, calm winter nights, indicating that some background factor shifted baseline UHI intensities throughout the year. We found that regional vegetation and snow cover conditions set distinct seasonal baselines for UHI intensity, with nighttime intensities averaging 4°C in summer and 1°C in winter. Synoptic and biophysical factors that vary on shorter time scales (e.g. wind, clouds, soil moisture, relative humidity) modified daily UHI intensity around those baselines by 1-3°C but were not the primary drivers of UHI seasonality, contrary to the most common hypotheses in the literature.

In addition to clarifying the drivers of UHI seasonality, our results have implications for the design and interpretation of UHI studies, particularly in how the timing of data collection might affect results and conclusions.