Forecasting seasonal to interannual variability in extreme sea levels

Abstract

Abstract Menendez, M., Mendez, F. J., and Losada, I. J. 2009. Forecasting seasonal to interannual variability in extreme sea levels. – ICES Journal of Marine Science, 66: 1490–1496. A statistical model to predict the probability of certain extreme sea levels occurring is presented. The model uses a time-dependent generalized extreme-value (GEV) distribution to fit monthly maxima series, and it is applied for a particular time-series record for the Atlantic Ocean (Newlyn, UK). The model permits the effects of seasonality, interannual variability, and secular trends to be identified and estimated in the probability distribution of extreme sea levels. These factors are parameterized as temporal functions (linear, quadratic, exponential, and periodic functions) or covariates (for instance, the North Atlantic Oscillation index), which automatically yield the best-fit model for the variability present in the data. A clear pattern of within-year variability and significant effects resulting from astronomical modulations (the nodal cycle and perigean tides) are detected. Modelling different time-scales helps to gain a better understanding of recent secular trends regarding extreme climate events, and it allows predictions to be made (for example, up to 2020) about the probability of the future occurrence of a particular sea level.