Revisited mixed extreme wave climate model for reanalysis data bases

Mixed extreme value models (Mínguez et al., Stoch Environ Res Risk Assess 27:757–768, 2013b ) have proved to be an appropriate tool for dealing with wave maxima because they take full advantage of upper tail information from both (1) hindcast or wave reanalysis and (2) instrumental records, which reduces the uncertainty on return level estimates. However, in order to characterize stochastically the differences between instrumental and reanalysis maxima, the method developed in Mínguez et al. (Stoch Environ Res Risk Assess 27:757–768, 2013b ) only uses information about annual maxima. This technical note revisits the MEV method so that those differences between instrumental and reanalysis maxima could be characterized using information on independent storm peaks, instead of annual extremes. This strategy increases the size of data sets during the estimation process, reducing uncertainty. The revisited mixed extreme value model is illustrated using data from the same location studied in Mínguez et al. (Stoch Environ Res Risk Assess 27:757–768, 2013b ), and results are compared.