We quantify the change in extreme high sea level (ESL) statistics in the German Bight under rising $${{\mathrm {CO}}_2}$$ concentrations by downscaling a large ensemble of global climate model simulations using the regionally coupled climate system model REMO-MPIOM. While the model setup combines a regionally high resolution with the benefits of a global ocean model, the large ensemble size of 32 members allows the estimation of high return levels with much lower uncertainty. We find that ESLs increase with atmospheric $${{\mathrm {CO}}_2}$$ levels, even without considering a rise in the background sea level (BSL). Local increases of up to 0.5 m are found along the western shorelines of Germany and Denmark for ESLs of 20–50 years return periods, while higher return levels remain subject to sampling uncertainty. This ESL response is related to a cascade of an enhanced large-scale activity along the North Atlantic storm belt to a subsequent local increase in predominantly westerly wind speed extremes, while storms of the major West-Northwest track type gain importance. The response is seasonally opposite: summer ESLs and the strength of its drivers decrease in magnitude, contrasting the response of the higher winter ESLs, which governs the annual response. These results have important implications for coastal protection. ESLs do not only scale with the expected BSL rise, but become even more frequent, as preindustrial 50-year return levels could be expected to occur almost every year by the end of the century. The magnitude of the relative change in ESL statistics is hereby up to half of the expected rise in BSL, depending on the location. Changes in the highest extremes are subject to large multidecadal variations and remain uncertain, thus potentially demanding even further safety measures.

Abstract. We investigate the long-term variability of extreme high sea levels (ESL) in the southern German Bight and associated large-scale forcing mechanisms in the climate system using simulations covering the last 1000 years. To this end, global MPI-ESM simulations from the PMIP3 past1000 project are dynamically scaled-down with a regionally coupled climate system model focusing on the North Sea. We find that the statistics of simulated ESL compare well with observations from the tide gauge record at Cuxhaven but show large variations on interannual to centennial timescales. ESL arise independent of preferred systematic oscillations and are to a large extent decoupled from variations of the background sea level (BSL). Large scale circulation regimes associated with periods of high ESL are regionally consistent and similar to those associated with elevated BSL, but the location of the respective centers of action of the governing sea level pressure (SLP) dipole differs. While BSL variations correlate well with the wintertime North Atlantic Oscillation (NAO), ESL variations are rather associated with a dipole between northeastern Scandinavia and the Gulf of Biscay, leading to a stronger local north-westerly wind component in the North Sea. Potential links with solar or volcanic forcing are masked due to the high ESL variability. The high internal variability stresses the irreducible uncertainties related to traditional extreme value estimates based on shorter subsets which fail to account for long-term variations. Existing estimates of future changes in ESL may be dominated by natural variability rather than climate change signals, thus requiring larger ensemble simulations to assess future flood risks.

formal assessment Does the paper address relevant scientific questions within the scope of OS? yes Does the paper present novel concepts, ideas, tools, or data? yes Are substantial conclusions reached? yes Are the scientific methods and assumptions valid and clearly outlined? yes Are the results sufficient to support the interpretations and conclusions? yes Is the description of experiments and calculations sufficiently complete and precise to allow their reproduction by fellow scientists (traceability of results)? yes