*3.5. Wave Storm Seasonality*

To assess the existence of seasonal variations in the timing of wave storms throughout the annual period, the day of the calendar year on which the maximum significant wave height of a given storm occurred has been converted to an angular value, *θ*, assuming that the number of days per year is 365. Thus, in leap years, the data corresponding to 29 February have been removed [11].

To examine whether storms in a given region exhibit a seasonal pattern, it is necessary to know whether it is statistically possible to accept that the time of occurrence of wave storms throughout the year is uniformly distributed. The acceptance or rejection of the uniformity hypothesis is assessed in this study through the use of the Rayleigh and Kuiper tests, by considering the storm peaks' timing throughout the year as a circular variable. The selection of these two tests from among the multitude of existing alternatives is due to the fact that the Rayleigh test is powerful only when it is possible to assume that the population distribution has only one mode, while the Kuiper test is specifically indicated in the case of multimodal distributions. More detailed information on these and other uniformity tests can be found in [11,29] and references therein.
