Sea-ice concentration in the Arctic and the North-Atlantic part of the sub-Arctic has undergone pronounced trends in recent decades. Although driven in large part by changes in atmospheric flow, which has shown pronounced trends toward the positive phase of the North-Atlantic/Arctic Oscillation (NAO/AO), the trend in sea ice exerts a significant feedback upon the atmospheric circulation. A modeling study where we examined the nature of this feedback and the feedback due to multidecadal trends in sea surface temperature (SST) in the extratropical North Atlantic in winter appeared recently in the Journal of Climate (March, 2004). The feedback due to multidecadal SST trends was weak and positive, whereas the feedback due to sea-ice trends was substantially stronger and negative. This feedback was highly sensitive to the location of the area of forcing, with the most sensitive area immediately East of Greenland. We are in the process of extending that part of our study to include a more detailed evaluation of the nature of the significant feedback due to sea-ice trends by representing fractional sea-ice cover over a gridbox. Surface characteristics that are modified by sea-ice concentration anomalies such as sensible and latent heat flux as well as longwave radiative cooling and albedo are allowed to change continuously.

Another important aspect that we address here and that was neglected in the previous study is to examine the nature of the subseasonal (monthly) feedback, especially late winter into spring and summer. Climate patterns other than the NAO/AO, such as the EA pattern (sometimes referred to as the BO (Barents Oscillation) in recent literature) become important outside the winter season. Our working hypothesis is that internal atmospheric variability is still prominently projected onto the dominant climate pattern so that the atmospheric response may still be decomposed into a global and a local part. We use the Community Climate System Model (CCSM) in our studies (http://www.ccsm.ucar.edu/). The latest version of the atmospheric model is CAM3, which was released on June 23 2004.

Thirdly, we plan to examine the nature of the feedback due to the forcing immediately downstream of Greenland in a limited area model, NCAR's MM5, as well as a higher resolution version of CAM3. The MM5 has been used by our collaborator in Iceland (Olafsson) specifically to examine effects of the orographic barrier of Greenland on cyclogenesis in the area of the North Atlantic, downstream of Greenland. The higher resolution is required in order to resolve the topography of Greenland.

Lastly, we will address the validity of using only the last 40 years in evaluating sea-ice concentration trends in light of the low frequency signal in Arctic climate trends, especially in the North Atlantic sector. Of course, sea-ice extent is known the best since satellite measurements became routine, approximately 25 years ago. Reliable measurements certainly go further back, especially in the North Atlantic sector, and even further back than the 40 years that we considered earlier. We intend to run sensitivity experiments using longer records, and decadal trends over decades when the trend was different to what it has been over the last four decades.