The changing atmospheric water cycle in Polar Regions in a warmer climate

We have examined the atmospheric water cycle of both Polar Regions, polewards of 60 degrees N and 60 degrees S, using the ERA-Interim reanalysis and high-resolution simulations with the ECHAM5 model for both the present and future climate based on the IPCC, A1B scenario. The annual precipitation in ERA-Interim amounts to similar to 17000 km(3) and is more or less the same in the Arctic and the Antarctic, but it is composed differently. In the Arctic the annual evaporation is similar to 8000 km(3) but similar to 3000 km(3) less in the Antarctica where the net horizontal transport is correspondingly larger. The net water transport of the model is more intense than in ERA-Interim, in the Arctic the difference is 2.5% and in the Antarctic it is 6.2%. Precipitation and net horizontal transport in the Arctic has a maximum in August and September. Evaporation peaks in June and July. The seasonal cycle is similar in Antarctica with the highest precipitation in the austral autumn. The largest net transport occurs at the end of the major extra-tropical storm tracks in the Northern Hemisphere such as the eastern Pacific and eastern north Atlantic. The variability of the model is virtually identical to that of the re-analysis and there are no changes in variability between the present climate and the climate at the end of the 21st century when normalized with the higher level of moisture. The changes from year to year are substantial with the 20- and 30-year records being generally too short to identify robust trends in the hydrological cycle. In the A1B climate scenario the strength of the water cycle increases by some 25% in the Arctic and by 19% in the Antarctica, as measured by annual precipitation. The increase in the net horizontal transport is 29% and 22%, respectively, and the increase in evaporation correspondingly less. The net transport follows closely the Clausius-Clapeyron relation. There is a minor change in the annual cycle of the Arctic atmospheric water cycle with the maximum transport and precipitation occurring later in the year. There is a small imbalance of some 4-6% between the net transport and precipitation minus evaporation. We suggest that this is mainly due to the fact that the transport is calculated from instantaneous six hourly data while precipitation and evaporation is accumulated over a 6-h period. The residual difference is proportionally similar for all experiments and hardly varies from year to year.