Moritz Mathis

Assessment of a zoomed global model for the North Sea by comparison with a conventional nested regional model

Assessment of marine downscaling of global model simulations to the regional scale is a prerequisite for understanding ocean feedback to the atmosphere in regional climate downscaling. Major difficulties arise from the coarse grid resolution of global models, which cannot provide sufficiently accurate boundary values for the regional model. In this study, we first setup a stretched global model (MPIOM) to focus on the North Sea by shifting poles. Second, a regional model (HAMSOM) was performed with higher resolution, while the open boundary values were provided by the stretched global model. In general, the sea surface temperatures (SSTs) in the two experiments are similar. Major SST differences are found in coastal regions (root mean square difference of SST is reaching up to 2°C). The higher sea surface salinity in coastal regions in the global model indicates the general limitation of this global model and its configuration (surface layer thickness is 16 m). By comparison, the advantage of the absence of open lateral boundaries in the global model can be demonstrated, in particular for the transition region between the North Sea and Baltic Sea. On long timescales, the North Atlantic Current (NAC) inflow through the northern boundary correlates well between both model simulations (R~0.9). After downscaling with HAMSOM, the NAC inflow through the northern boundary decreases by ~10%, but the circulation in the Skagerrak is stronger in HAMSOM. The circulation patterns of both models are similar in the northern North Sea. The comparison suggests that the stretched global model system is a suitable tool for long-term free climate model simulations, and the only limitations occur in coastal regions. Regarding the regional studies focusing on the coastal zone, nested regional model can be a helpful alternative.

The subpolar gyre regulates silicate concentrations in the North Atlantic

The North Atlantic is characterized by diatom-dominated spring blooms that results in significant transfer of carbon to higher trophic levels and the deep ocean. These blooms are terminated by limiting silicate concentrations in summer. Numerous regional studies have demonstrated phytoplankton community shifts to lightly-silicified diatoms and non-silicifying plankton at the onset of silicate limitation. However, to understand basin-scale patterns in ecosystem and climate dynamics, nutrient inventories must be examined over sufficient temporal and spatial scales. Here we show, from a new comprehensive compilation of data from the subpolar Atlantic Ocean, clear evidence of a marked pre-bloom silicate decline of 1.5–2 µM throughout the winter mixed layer during the last 25 years. This silicate decrease is primarily attributed to natural multi-decadal variability through decreased winter convection depths since the mid-1990s, a weakening and retraction of the subpolar gyre and an associated increased influence of nutrient-poor water of subtropical origin. Reduced Arctic silicate import and the projected hemispheric-scale climate change-induced weakening of vertical mixing may have acted to amplify the recent decline. These marked fluctuations in pre-bloom silicate inventories will likely have important consequences for the spatial and temporal extent of diatom blooms, thus impacting ecosystem productivity and ocean-atmosphere climate dynamics.

Projected Change—North Sea

Increasing numbers of regional climate change scenario assessments have become available for the North Sea. A critical review of the regional studies has helped identify robust changes, challenges, uncertainties and specific recommendations for future research. Coherent findings from the climate change impact studies reviewed in this chapter include overall increases in sea level and ocean temperature, a freshening of the North Sea, an increase in ocean acidification and a decrease in primary production. However, findings from multi-model ensembles show the amplitude and spatial pattern of the projected changes in sea level, temperature, salinity and primary production are not consistent among the various regional projections and remain uncertain. Different approaches are used to downscale global climate change impacts, each with advantages and disadvantages. Regardless of the downscaling method employed, the regional studies are ultimately affected by the forcing global climate models. Projecting regional climate change impacts on biogeochemistry and primary production is currently limited by a lack of consistent downscaling approaches for marine and terrestrial impacts. Substantial natural variability in the North Sea region from annual to multi-decadal time scales is a particular challenge for projecting regional climate change impacts. Natural variability dominates long-term trends in wind fields and strongly wind-influenced characteristics like local sea level, storm surges, surface waves, circulation and local transport pattern. Multi-decadal variations bias changes projected for 20- or 30-year time slices. Disentangling natural variations and regional climate change impacts is a remaining challenge for the North Sea and reliable predictions concerning strongly wind-influenced characteristics are impossible.

Which complexity of regional climate system models is essential for downscaling anthropogenic climate change in the Northwest European Shelf

Climate change impact studies for the Northwest European Shelf (NWES) make use of various dynamical downscaling strategies in the experimental setup of regional ocean circulation models. Projected change signals from coupled and uncoupled downscalings with different domain sizes and forcing global and regional models show substantial uncertainty. In this paper, we investigate influences of the downscaling strategy on projected changes in the physical and biogeochemical conditions of the NWES. Our results indicate that uncertainties due to different downscaling strategies are similar to uncertainties due to the choice of the parent global model and the downscaling regional model. Downscaled change signals reveal to depend stronger on the downscaling strategy than on the model skills in simulating present-day conditions. Uncoupled downscalings of sea surface temperature (SST) changes are found to be tightly constrained by the atmospheric forcing. The incorporation of coupled air–sea interaction, by contrast, allows the regional model system to develop independently. Changes in salinity show a higher sensitivity to open lateral boundary conditions and river runoff than to coupled or uncoupled atmospheric forcings. Dependencies on the downscaling strategy for changes in SST, salinity, stratification and circulation collectively affect changes in nutrient import and biological primary production.