Examining extreme sea levels for the support of nuclear power plant safety in Finland

Urbanized low-lying coastal areas become evidently increasingly vulnerable in the future climate, where higher mean sea level and more powerful storm surges are expected. Safe planning and operation particularly of critical coastal infrastructure, such as nuclear power plants, requires careful evaluation of external events due to extreme weather and sea level conditions. Finland is located in the northeast part of the Baltic Sea, where sea level variations are driven by short-lasting phenomena (storm surges, internal oscillations in the Baltic Sea, tides), long-term mean sea level change (global mean sea level, the post-glacial land uplift and the Baltic Sea water balance), and wind waves. In Finland, there are altogether five nuclear power plant units along the coast and approximately one third of electricity production in the country is grounded on nuclear energy.

The Finnish Meteorological Institute has studied weather, climate and sea level hazards potentially posing risks to nuclear power plants since 2007 in several research projects. In this presentation, we will introduce some preliminary results of studies on extreme sea levels on the Finnish coast, which are conducted in the PREDICT project (https://en.ilmatieteenlaitos.fi/predict). Their focus is on the short-term sea level variations, which might be several meters in the Baltic Sea, even if the tides in the region are mostly negligible.

In the first study, Bayesian hierarchical modelling is used to estimate return levels of annual sea level maximum on the Finnish coast and non-stationarity in the related sea level extremes. Our model setup enables sharing information on sea level extremes between the neighboring tide gauge stations. Additionally, it accounts temporal variations in the distribution parameters by incorporating climate indices such as North Atlantic oscillation (NAO) in the model. Preliminary results suggest that hierarchical approach reduces the range of uncertainty in the estimated parameters.

The second study tackles a question “What is the most severe flooding that could occur in the Baltic Sea coast in the present climate if the weather conditions are optimal?”. In this study, effects of low-pressure intensity, speed, direction and point of origin on the sea level extremes is examined by making simulations with numerical sea level model combined with the synthetic cyclones for the Finnish tide gauge locations. Tentative results indicate that the highest sea level extremes on the Finnish coast are caused by large and slowly propagating wind storms.