Abstract
This paper examines long-term change in the interannual variability in surface air temperature and its cause by using monthly data generated in climate change experiments (1851-2100) performed using two models. Regions north of 20°N are analyzed. Anomalies are defined as high-pass filtered values with a cutoff period of 30 years. Interannual variabilities of anomalies are expressed as the root mean square value for a 30 year period.
Before global warming (around year 1900), the interannual variability in temperature is large over sea areas north of 50°N. This is because a large gradient of sea ice concentration brings about a large temperature gradient there. The interannual variability decreases generally in a cold season at high latitudes with global warming. In contrast to the general decrease of the interannual variability, there are some regions where it increases locally north of regions that show significant decrease around the Arctic. It can be understood that these phenomena are brought about by the decrease in the sea ice concentration gradient (i.e., the decrease in the temperature gradient) in the southern part of the polar region and the increase in the sea ice concentration gradient (i.e., the increase in the temperature gradient) in the northern part, due to the northward shift of sea ice edges. However, the degree of the decrease of the interannual variability is larger than that of the increase. This is because the global decrease of the temperature gradient strengthens the decrease of the interannual variability and weakens the increase. In addition, even over regions without sea ice, there are some areas where the interannual variability decreases. This is also because the temperature gradient decreases globally.
