U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications

Intrinsic versus forced variation in coupled climate model simulations over the Arctic during the Twentieth Century

Muyin Wang

Joint Institute for the Study of the Atmosphere and Ocean, University of Washington, Seattle, Washington

James E. Overland

National Oceanic and Atmospheric Administration/Pacific Marine Environmental Laboratory, Seattle, Washington

Vladimir Kattsov

Voeikov Main Geophysical Observatory, St. Petersburg, Russia

John E. Walsh and Xiangdong Zhang

International Arctic Research Center, University of Alaska Fairbanks, Fairbanks, Alaska

Tatyana Pavlova

Voeikov Main Geophysical Observatory, St. Petersburg, Russia

J. Climate, 20(6), 1093–1107 (2007).
Copyright 2007 American Meteorological Society. Further electronic distribution is not allowed.

4. The spatial distribution of midcentury warm anomalies

Based on observations from 59 Arctic stations, Overland et al. (2004) found that warm anomalies in the midcentury were regional and episodic, that is, interannual to multiyear. An example of the spatial distribution of 5-yr-averaged winter temperature anomalies from CRUTS2.0 (Fig. 9, top left) shows that in the mid-1930s the warm anomalies were regional in extent: Eurasia and Greenland were dominated by warm anomalies with largest amplitude in northern Scandinavia, while North America was occupied by cold anomalies. Around the 1940s (Fig. 9, bottom left), warm anomalies were found with the largest amplitudes over central Alaska and east Siberia. During both periods there was an out of phase (seesaw) pattern between the eastern and western Arctic.


FIG. 9. Spatial distribution of LSAT anomalies based on (left) CRUTS2.0 and (right) one realization each from the eight models that pass the 2/3CRU criterion. All patterns have a 5-yr running mean applied. The years selected are around maxima of the warm anomalies during 1911–60 for both the observation and models. Contour interval is 0.5°C.

The spatial patterns from one realization for each of the eight models identified by the 2/3CRU criterion are shown in Fig. 9. All eight display warm anomalies over Alaska, except ECHAM5/MPI-OM (bottom-right panel). Five models (CSIRO-Mk3.0, ECHAM5/MPIOM, ECHO-G, INM-CM3.0, and GFDL-CM2.0) produced the seesaw pattern. CCSM3 and both GFDL models produced a pattern similar to the mid-1930s of CRUTS2.0 (top-left panel in Fig. 9) with warm anomalies over the central Eurasian continent and cold anomalies over the central North American continent and along the east coast of Greenland. The ECHAM5/MPI-OM model also produced a similar feature; however, the center is over the continent instead of over Scandinavia. The pattern produced by CSIRO-Mk3.0 is similar to the 1940s observations (bottom-left panel in Fig. 9). The ECHO-G model displays the seesaw pattern, with the amplitudes of the warm anomalies weaker than the observed, and the position of the cold anomaly center shifted to the east. The PCM model has wavenumber-1 pattern, but the region of the cold anomalies is smaller than the observed. The contrast of observed warm anomalies in 60°–70°N and cold anomalies between 50° and 60°N over Eurasia is not seen in these models. Even though models agree on domain-averaged anomalies, their spatial distributions differ significantly, which also suggests the importance of regional intrinsic variability due to shifts in atmospheric circulation patterns.

Return to previous section or go to next section

PMEL Outstanding Papers

PMEL Publications Search

PMEL Homepage