Comments on “Point-in-time and extreme-value probability simulation technique for engineering design”, by R. Minguez, Y. Guanche and F. Méndez [Structural Safety Vol. 41 pp. 29–36, 2013]

Mínguez et al. (2013) [1] proposed a method, (a) to characterize and to combine graphically both the point-in-time and extreme-value probability distribution, (b) to apply Monte Carlo simulation technique to reproduce both regimes, and (3) to use their distributions in the frame of first order reliability methods (FORM). This discussion is to comment on the limited validity of the proposed method. Firstly, the method should formally include the extremal index in the relationships between the distribution functions. Moreover, it is only valid if the data are independents and there is no autocorrelation or it is negative. To provide additional information related with the discussion, two new set of data are worked out: (1) significant wave height obtained in the Gulf of Cádiz (hindcast data, WANA 1952046) with extremal index 0.32, and (2) a simulated series with an ARMA model, autoregressive coefficient 0.95, zero moving average coefficient and an extremal index equal to 0.3.