Influence of Atmospheric Circulation on the Baltic Sea Level Rise under the RCP8.5 Scenario over the 21st Century

Sitar Karabil
2017 Climate  
This study aims to estimate the influence of atmospheric circulation modes on future Baltic Sea level rise under the Representative Concentration Pathway 8.5 (RCP8.5) climate scenario for the period 2006-2100. For this estimation, the connection between the sea level variations in two selected representative locations-Stockholm and Warnemünde, and two atmospheric indices-the Baltic Sea and North Sea Oscillation (BANOS) index and the North Atlantic Oscillation (NAO) index is statistically
more » ... tatistically analysed. Correlations of winter means between atmospheric indices, BANOS and NAO, and tide gauges are measured as 0.85 and 0.55 for Stockholm, and 0.55 and 0.17 for Warnemünde over the period 1900-2013. Assuming that the established connection remains unchanged, the influence of atmospheric circulation modes on future Baltic Sea level rise is estimated from the projections of atmospheric indices, which are constructed from the SLP outputs of climate models participating in phase 5 of the Coupled Model Intercomparison Project (CMIP5) under the RCP8.5 scenario. The main conclusion is that the contribution of those atmospheric modes to the Baltic Sea level rise is likely to remain small over the 21st century. Additionally, corresponding trend estimations of model realizations indicate the large influence of the internal climatic variability of the CMIP5 models on those future trends. One of the most important findings of this study is that anthropogenic forcing does not play a key role in the evolution of these atmospheric indices. studies focused on revealing the impact of the individual driving factors on sea-level variability [2] [3] [4] . For instance, Grinsted et al. [5] explored the sea-level rise projections in Northern Europe including the Baltic Sea under the Representative Concentration Pathway 8.5 (RCP8.5) scenario. Considering the regional fingerprints of some major sea level components of the global sea level budget like ocean thermal expansion, melting/dynamics of glaciers, ice loss from the Greenland and Antarctic ice sheets, and changes in land water storage, their best estimation of relative sea-level rise is 0.25 m for Stockholm under the RCP8.5 scenario over the 21st century. However, projections of global mean sea level rise for the RCP8.5 scenario are in the range between 0.52 and 0.98 m over the 21st century [1]. Besides, it is known that the most important factors causing sea level variations in the Baltic Sea from interannual to decadal time scales are connected to the variability of atmospheric circulation [6] [7] [8] . Therefore, modifications in the atmospheric conditions are important driving factors for understanding the possible sea-level rise in the near future, especially in the regions like the Baltic Sea, where the variability of sea level is quite sensitive to atmosphere driven boundary conditions, such as wind, air pressure, air temperature, and freshwater flux [9] [10] [11] . Since the NAO can carry information about large scale changes in the pressure field, wind field, heat fluxes, and freshwater fluxes, there are several studies investigating the link between the NAO and sea-level variability in the Baltic Sea [8, 12] . Two main conclusions on that link are that the NAO has a heterogeneous effect on the sea-level variability in the spatial domain and that this link is also not stable in time [11] . By analyzing the winter monthly means of sea level and SLP patterns, Heyen et al. [13] established a statistical link between the leading vectors of those variables which are described by a canonical correlation analysis (CCA). They detected a strong connection between an individual mode of large-scale SLP and the sea level variation pattern in the Baltic Sea region. They concluded that wind stress over the transition area (covering the Skagerrak, the Kattegat, and the Danish Straits), inverse barometric effect (IBE), and sea level variation in the German Bight are associated with the leading sea level CCA vector in the Baltic Sea. However, they did not find any considerable contribution of the precipitation to that leading mode of sea level variation. More recently, Hünicke and Zorita [8] investigated the role of atmospheric factors such as three-to-five leading vectors of large-scale SLP, sea surface temperature, and precipitation in the Baltic Sea level variations by establishing a statistical model between the individual tide gauges and those variables (stepwise multivariate regression) on interannual time scales. By filtering time series with 11-year running means, they first showed that a non-negligible amount of sea-level variability cannot be linearly explained by the SLP field. Following this preliminary result, they detected spatially heterogeneous contributions of precipitation and temperature to Baltic Sea level variability. Further analysis is applied by Hünicke et al. [14] on a decadal time series of winter means. They confirmed that the impact of atmospheric forcing on decadal sea-level variability is also geographically heterogeneous in the Baltic Sea region. Hünicke et al. [14] found that sea-level variability in the southern Baltic can be better explained by spatially averaged precipitation; besides, atmospheric circulation patterns are dominant factors in modifying sea level variations in the central, eastern, and northern Baltic Sea on decadal time scales. However, temperature does not play a role in causing variation in the Baltic Sea level on decadal time scales, in contrast to the results of Hünicke and Zorita [8] . It can be the case that the strength of the relation between sea level and a climatic factor becomes important due to the fact that the mean state of the system can be changed slowly on decadal time scales, as that was likely the case for the different impact of temperature on sea level between the studies Hünicke and Zorita [8] and Hünicke et al. [14] . Those results of Hünicke and Zorita [8] and Hünicke et al. [14] also seem to be contradictive in comparison to the study of Heyen et al. [13] , which did not detect a considerable effect of precipitation on sea level variability in the Baltic Sea. This discrepancy may mainly arise from the fact that Heyen et al. [13] used a CCA-method by using principal components of sea level and of SLP field, which cannot capture the effects that are not large-scale and spatially incoherent over the whole Baltic Sea region.
doi:10.3390/cli5030071 fatcat:asxgdw73fras5kufqwfmrbmazq