Understanding temperature responses to climate change : distributions, extremes, and seasonal cycle

Understanding future changes in temperature variability and extremes is an important scientific challenge with societal impacts. Previous studies have developed theories based on convective coupling to interpret changes in extreme temperatures over land. A novel theoretical framework based on local convective coupling is developed that highlights eleven physical mechanisms with the potential to drive differential warming. In climate-warming simulations with both idealised and more realistic models, the tropical land summer temperature distribution broadens because hot days are dry. Over tropical oceans, the summer temperature distribution narrows with warming because the aloft lapse rate increases relatively more on cold days. The spatial distribution of the projected warming of the hottest days can be understood with the same framework. The projected near-surface relative humidity change in response to warming is the dominant mechanism responsible for spatial variability; in particular, relative drying over South America and much of Europe drives their amplified warming. However, the framework is not appropriate to apply to eastern Canada, where our simulations project the greatest warming of hot days, with a surface-based perspective more suitable. This surface perspective is used in a different context to quantify the meridional variation in the date and temperature of the hottest and coldest days of the year in idealised simulations. Previous studies state the importance of the polar night character of solar radiation at high latitudes in pushing the coldest day later and pulling the hottest day earlier in the year. We explicitly quantify this effect and also show that it acts to increase the temperature of both the hottest and coldest days. The role of other variables such as near-surface wind speed and relative humidity on the hottest and coldest days is also discussed in a quantitative manner.