A detailed study of rainfall in the Roman area in the decade 1992–2001

A study of the rainfall regime in the Roman area over the decade 1992-2001 has been undertaken on using tipping pluviometers data, coming from 23 climatic stations located in Rome and in its surroundings. The time response of the instruments and the automatic acquisition system ensure a resolution of less than 1 minute, thus offering the possibility of an accurate evaluation of intense and extreme events. The mean yearly rainfall over the whole decade has been determined for each station, obtaining values between 682 and 870mm/year, with a geographical average of 771 and a standard deviation of 47. A study of the rainfall distribution within the 48 half-hours of the day has been carried out in order to ascertain whether preferred times for rain events exist. The analysis has evidenced that this is the case, with high rainfall rates mostly occurring in the late morning and low rates in the late night. Typical values of the maximum 30-min rainfall ever recorded at any given station oscillate between 25 and almost 60 mm. A separate analysis of rainy and dry days has been carried out on studying the statistics of the time delays between two successive tips of the pluviometer. This allowed a characterization of the intense rains as well as of the droughts: the resulting histograms show the existence of a bimodal distribution explained in terms of two kinds of rain events, intense summer showers and drizzles distributed over the rest of the year. As for the droughts, the longest durations appear to range from one to about seven months. On confining the analysis to the rainy days only, the rain intensity data for each station has been plotted and fitted with a Weibull distribution. The corresponding Weibull parameters, while gathering around common mean values, do not show any recognizable pattern when regressed, for instance, versus the altitude of the station or the distance from the coastline. Last, the likelihood that a day of the year, taken at random, be a rainy day or not has been computed for each station yielding probability values ranging from 0.18 to 0.22