Advancing actionable knowledge on sealevel extremes through an ocean modelling framework

The consequences of climate change have begun to manifest globally rather than just in single regions. The risks associated with the combination of coastal exploitation and rising sea levels are becoming more apparent. The Intergovernmental Panel on Climate Change (IPCC) irrefutably links anthropogenic greenhouse gas emissions to climate change and thus the trajectories for sealevel rise projected for the current mitigation strategies (MassonDelmotte et al., 2021).

One obvious question emerges for future extreme sea levels; how well do we know the baseline for change regarding the most extreme sea levels? It can be necessary to secure societal functions against events that only occur once every thousand years or even more rarely. Compared to these timescales, even the time spans covered by our longest systematic observational records are fleeting. Statistical extrapolations of observed extremes have limited ability to provide us with the knowledge we require regarding lowprobability events. Such estimates will reflect the variability based on spatially heterogeneous “snapshots” in time (about a hundred years). Models of the ocean and atmosphere have helped alleviate this problem. They complement the observational record with reconstructions of the past and projections for the future. Essentially, models can extend our observational horizon and serve as laboratories to explore the drivers and mechanisms of extreme sealevel events.

The most notable extreme sealevel uncertainties are associated with events that could be deemed implausible to occur due to exceedingly low probability, but where the potential consequences are so catastrophic that the risks cannot be neglected (Chen et al., 2021). Such lowprobability events might have occurred once or not at all within our observational records. It is uncertain whether climate model simulations can simulate such events as they are constrained by our current understanding and representation of the underlying physics.

This thesis aims to improve our understanding of extreme sealevel events by probing the boundaries of what is physically plausible. Here, a highlyskilled ocean model combined with various types of atmospheric forcing is used as a laboratory for sealevel extremes in the seas around Denmark.

The experiments conducted throughout this project established the ocean model’s usefulness for such investigations; its ability to reproduce extreme sealevel events proficiently along the complex coastlines of Denmark is ascertained in the first of the studies presented in this thesis. The first study also explores ways to deal with deficiencies in the data used to drive the ocean model. These results are particularly useful for studies investigating alternative scenarios of the past by modifying the input data. The second study presented here provides systematic mappings of extreme sea levels depending on the wind direction. For this mapping, hazardous wind directions and relevant time scales were quantified and intuitively visualized for Denmark, and the analysis can be extended for any location within the model domain. The premise for this study is simplistic, synthetic wind fields that allow us to assess all possible wind directions. The synthetic wind makes up for the lack of historic events for some wind directions.

On time scales of decades or more, sealevel rise will cause presentday extremes to occur more often, but in some areas counteracting effects such as land rise can partly reduce this effect. The relative rates have implications for extreme sea levels since mean sealevel rise increases the likelihood of flooding. With this in mind, the third study of this thesis quantifies future mean sealevel rise for coastal stretches around Denmark.

Lastly, the fourth study investigates the role of antecedent conditions for extreme storm surges in the Baltic Sea. The adverse effects of natural hazards are dependent on the conditions before the hazard itself. We explore this by changing the antecedent conditions, with the outcome that a previously unprecedented storm surge event becomes even worse.

Overall, the findings of this thesis may facilitate disaster risk management and serve as a steppingstone towards quantifying risk levels for lowprobability, highimpact extreme sealevel events. Knowledge of such events is urgently needed for the prevention and mitigation of dangerous coastal flooding.