A new method of identifying an appropriate distance between independent extreme annual rain events for a 5-min time resolution precipitation data network.

This paper deals with precipitation events, particularly the problem of recognising events individually, of distinguishing one local meteorological/precipitation event from another. The final aim was to arrive at a more appropriate conceptualization of "distance" between individual annual events. A criterion for identifying annual precipitation events, inside a 5 min-timestep series of pluviometric data, was proposed. Using this numerical criterion, a series of annual distances between events were obtained and compared with those obtained by other two known and applied statistical methods. This comparison would be important for linking the so-called statistical methods to the real data of the series, analysed through a numerical process. This would make it possible to provide a kind of "measure" for the understanding of real data via a statistical model. Results showed that distances between events, obtained through numerical analyses, are almost always different from those obtained by statistical methods. The supposed Poisson-process, generally matched to, and widely applied in the field of rainfall data, showed some contradiction with the analysis conducted in this paper. The so-called Lag method shows a numerical method type distribution, but produces some important internal deviations from numerical results. It seems that the method proposed in this paper can reproduce the real distances between events more realistically compared to the statistical methods considered here. Exponential distribution does not seem highly performing, at least in the case of the annual analysis, with series of a 5-min time-step resolution, obtained from rain gauges installed in the Veneto Region of Italy. It is believed that the numerical criterion presented in this paper can be considered adequate for determining the distances between events, within an annual time-window. [ABSTRACT FROM AUTHOR]