We know that the physical ocean conditions are of importance for the marine ecosystems and thereby also fishery resources. The natural variability is large, and on short time scales (less than 15 years) this is more important than the slower global warming. However, on longer times scales (more than 30 years), global warming will in general dominate natural variability.
The global climate models are large and complex models that include atmosphere, ocean and sea ice, and long simulations demand large amounts of computing resources. Even with increasing computing power, these models have poor horizontal resolution in our regions. These models are relatively good in the large open basins, but less accurate in offshore seas such as the North Sea and the Barents Sea. Furthermore, the global models often have issues with sea ice processes. This is of particular importance in the Barents Sea which is the area in the Arctic with the greatest variability in the sea ice distribution.
A way to improve this is by downscaling, where we run our regional ocean model with forcing from one or more global climate models. Among other things, we have done such downscalings for various emission scenarios in an area that covers the North Atlantic, the Nordic Seas, the Barents Sea, and the Arctic. The figure shows the trend in March sea surface temperature in the SSP5-8.5 scenario from 2015 to 2100. This scenario gives a warming of more than 4oC along the Polar Front and in the eastern part of the Barents Sea, in the Greenland Sea and north of Iceland and Svalbard, and even a cooling east of Iceland. The animation shows temperature anomalies relative the mean value for the same period and the same scenario. The large year-to-year variability illustrates the natural variability on top of the trend shown in the figure. Though, it has to be taken into consideration that the uncertainty is great, and that the results vary both between the different global models and the different scenarios.
Published: 19.03.2019 Updated: 12.04.2021