Global prediction of cumulative rainfall statistics from the simple knowledge of the yearly rain amount

An upgraded version of a one-parameter model for the prediction of the 1-minute Cumulative Distribution Function of rain rate (P(R)) is presented and tested. The model retains the analytical formulation proposed by the ITU-R Rec. P.837-5, but two out of its three input parameters, the 6-hour rainy period probability, P_r6, and the convective-over-total precipitation ratio, beta, are replaced by their average values pertinent to the meteorologically homogeneous areas identified in the study. The third parameter, the yearly rain amount, M_t, is the only site specific input. Tests have shown the ability of this simple model to reproduce measured P(R)s with very good approximation.