Abstract
Over the past few decades, heat-island related temperature increases in Phoenix, Arizona have been similar to the temperature increases predicted in a number of greenhouse simulation experiments. In this investigation, we use the Phoenix climate record to assess how increasing summertime mean temperatures are related to changes in the extreme maximum and minimum temperatures. Generally, rising mean temperatures are associated with substantial changes in the occurrence of extreme minimum temperatures (e.g., fewer days of extreme low minimum temperatures and more days of extreme high minimum temperatures). However, while the rising mean temperatures strongly influence the occurrence of moderately high maximum temperatures, they are weakly associated with the occurrence of extreme maximum temperatures. The results suggest that considerable caution should be used in predicting the occurrence of extreme temperatures from projected increases in mean temperature levels.
