Bayesian hierarchical modelling of sea-level extremes in the Finnish coastal region

Abstract. Occurrence probabilities of extreme sea levels required in coastal planning, e.g. for calculating design floods, have been traditionally estimated individually at each tide-gauge location. However, these estimates include uncertainties, as sea-level observations typically only a small number cases such annual maxima. Moreover, exact information on extremes between the locations and incorporation dependencies adjacent stations is often lacking analysis. In this study, we use Bayesian hierarchical modelling to estimate return maxima short-term variations related storm surges Finnish region. We generalised value (GEV) distribution basis compare three model structures different complexity against tide-gauge-specific fits. The allow us share maximum neighbouring also provide natural way uncertainties theoretical estimates. results show that compared fits, models, which pool across tide gauges, narrower uncertainty ranges both posterior parameter corresponding most locations. shape GEV systematically negative indicates Weibull type behaviour along coast. allows calculate upper limit Depending gauge considered, median was 47–73 cm higher than highest observed level.