A refined analysis and simulation of the wind speed macro-meteorological components

The authors have recently developed a Monte Carlo simulation algorithm able to generate thousands of years of synthetic wind speed values. In this study, however, two points have not received adequate attention: the presence of underlying periodic deterministic components embedded into the time series of the velocity and the generation of time series having exactly the desired marginal distribution. In principle, both these issues may have relevant consequences on the distribution of extreme values. This paper investigates and discusses the presence of underlying deterministic components in the wind records, develops a refined simulation algorithm for non-Gaussian colored wind speed stationary processes, and finally inspects how much these advances affect the analysis of extreme values. An extensive simulation consisting of more than 12,740 years of synthetic wind speeds is performed, based on the synoptic wind climatology of the central part of Italy. The results demonstrate that the use of a refined simulation algorithm improves the quality of extreme values analysis. On the other hand, it is not easy to express a definite opinion on the random or deterministic nature of some periodic cycles; it is shown, however, that dealing with such cycles as random or deterministic has very limited consequences on extreme values analysis.