Scientists from the Bjerknes Centre and the Nansen Center in Bergen recently presented this finding in Geoscientific Model Development.
Climate models are the best tools available to test hypotheses about the factors causing climate change and to assess future Earth system developments. In this article the research team tested to what degree Bergen Climate Model is able to simulate important features of the observed climate over the last 150 years.
Important aspects of the simulated general atmosphere- and ocean circulation, in addition to sea-ice extent, have been evaluated against observed data. The model has been run for over 600 years, and is relatively stable over this period.
Except for the very high latitudes, the model is able to reproduce the observed large-scale circulation in the atmosphere quite realistically. In addition, important natural modes of climate variability, such as the North Atlantic Oscillation, is realistically simulated. Several improvements in the ocean model, have led to a more realistic meridional overturning circulation compared to earlier versions of the model. The simulated volume transports in and out of the Nordic Seas also compares favorably with observed estimates.
The simulated sea-ice extent in the Northern Hemisphere is in general agreement with observational data except for summer where the extent is somewhat underestimated. In the Southern Ocean, large negative biases are found in the simulated sea-ice extent. This is partly related to problems with the mixed layer parameterization, causing the mixed layer in the Southern Ocean to be too deep, which in turn makes it hard to maintain a realistic sea-ice cover here.
In summary, the model is able to simulate important features of the observed large-scale circulation in the atmosphere as well as in the ocean. However, the instrumental record covers only a small fraction of the possible range of climate variability. Therefore, comparison of modeled climate states with larger climate perturbations that occurred in the past, and can be reconstructed from geological archives, can provide further evaluation of the model.