Natural Climate Variability in the Arctic Atmosphere
Mean 500 hPa geopotential height (gpm) of 12 warm Januaries (left) and 12 cold Januaries (right) simulated with HIRHAM.
Background:
The Arctic climate in the past and present is characterized by a strong variability on annual, decadal and centennial time-scales. The causes of these fluctuations are difficult to detect, because the climate system is controlled by a combination of external forcings, anthropogenically induced changes as well as internal instabilities and feedback effects. As mechanisms for climate variability on time-scales of decades to centuries, nonlinear instability and feedback processes in the atmosphere itself and in the atmosphere-ocean-ice system are possible natural causes. For the last years it has been accepted that natural variability is also the key to progress on discussions of the detection of anthropogenic climate changes.
The KIHZ Project:
The project "Natural climate variations from 10,000 years to the present day" (Klima in historischen Zeiten, KIHZ) is a joint effort to analyze the dynamics of natural climate variability. Data from continental, marine and atmospheric archives are systematically combined with dynamic realistic climate modeling and data assimilation. The project thus comprises three main segments,
- analysis of geological archives
- evalution of existing paleoclimate data and time series
- global and regional climate modeling.
Time series of the winter (DJFM) index of the NAO from observations (top) and form the control run of the ECHO-G (bottom). The green bars represent selected NAO+ and NAO- periods for the regionalization of the Arctic climate.
Regional Climate Modeling
In a first study we analyzed a long-term simulation with a constant external forcing according to the present-day climate (control run) of the global coupled atmosphere-ocean circulation model ECHO-G (ECHAM4/HOPE-G) to find periods persisting over several years either with mainly warm or mainly cold conditions during the Arctic winter. Two warm and two cold periods, each of 6 years duration, were selected for the regionalization of the Arctic climate in January. The performed simulations with our regional atmospheric climate model HIRHAM show that a warm or cold Arctic winter climate is connected with two distinct circulation states of the Arctic atmosphere. These states are characterized by a different location and extension of the tropospheric vortex with a vortex center over the western Arctic in warm Januaries or over the eastern Arctic in cold Januaries, respectively (see Figure 1). More detailed information is given in:
- Dorn, W., K. Dethloff, A. Rinke, M. Botzet, 2000: Distinct circulation states of the Arctic atmosphere induced by natural climate variability, J. Geophys. Res., 105, 29659-29668.
One major source of natural climate variability in the Northern Hemisphere is the North Atlantic Oscillation (NAO). In order to analyse the influence of this mode on the Arctic climate, HIRHAM simulations have been carried out for several positive and negative periods of the NAO (labeled as NAO+ and NAO-). The lower and lateral boundary forcing for the HIRHAM were taken from different global data sets:
- Selected years of the 1960s and 1970s from the NCEP/NCAR reanalysis and of the 1980s and 1990s from the ECMWF reanalysis.
- Selected periods from the above mentioned control run of the ECHO-G.
Differences between NAO+ and NAO- phases of the mean wintertime (DJFM) precipitation (in millimeters) from HIRHAM simulations with ECHO-G boundary forcing.
The NAO has a strong influence on the climate conditions in the Atlantic-European sector of the Arctic, but also on the climate in the southern and western part of Greenland. Over North-Europe a warmer and wetter climate is associated with the positive phase of the NAO, while over South- and West-Greenland higher temperatures and more precipitation occur in the negative phase. Moreover, the precipitation distribution shows a different behavior with regard to the NAO in West- and East-Greenland (see Figure 3).
For a reconstruction of the NAO index by paleoclimate data as snow accumulation rates analyzed from Greenlandic ice cores, it seems to be very important to choose the right location for the corresponding ice core. However, the simulations show that ice cores from the summit of Greenland are not qualified for the reconstruction of the NAO, while ice cores from western Greenland might be much better for it.
More detailed information is given in:
- Dorn, W., 2001: Natürliche Klimavariationen der Arktis in einem regionalen hochauflösenden Atmosphärenmodell (Dissertation an der Universität Potsdam).