Thermal comfort is quantified in 15 regions using the wet-bulb globe temperature (WBGT), examining past and future rates of thresholds exceedance corresponding to moderate, high, and extreme heat (28, 32, and 35 °C, respectively). As recent changes to thermal comfort appear to be dominated by temperature and humidity, a WBGT approximation based only on these is used. A new homogenised dataset from 1973 to 2003 is developed which provides WBGT daily means, daily maximums averaged over 5-day periods, and the highest extreme for each 5-day period; recent trends are positive for all regions except northeast USA and northeast Australia. A simple model for predicting summertime threshold exceedance rates, with a fixed distribution of anomalies about the seasonal mean, is found to adequately predict changes for the above quantities given seasonal mean values. This model is used to predict the impact of regional 1–5 °C temperature increases on WBGT exceedance rates with no change in relative humidity. Results show that heat events may worsen as much, or more, in humid tropical and mid-latitude regions even if they warm less than the global average, due to greater absolute humidity increases. A further 2 °C warming from the present is sufficient to push peak WBGT above 35 °C, an extreme heat event, in all regions except the UK.
An ensemble of HadCM3 climate model simulations is used to investigate likely regional changes in mean summertime temperature, relative humidity and WBGT under an A1B scenario for the 2020s and 2050s. Unsurprisingly, simulated regional changes often depart significantly from the global average, and the impact of regional changes in relative humidity is not always negligible. Increases in WBGT are nonetheless expected in all regions, and are more predictable than increases in temperature at least in mid-latitude regions owing to the compensating effects of humidity. © 2010 Royal Meteorological Society and Crown Copyright