PSI - Issue 84

Michele Placido Antonio Gatto et al. / Procedia Structural Integrity 84 (2026) 111–118

114

Fig. 2. Example of the derivation of the D2 distribution: (a) cumulative rainfall over calendar months; (b) mean daily background rainfall; (c) application to the 20-year reference period.

Fig. 3. Example of the derivation of the D3 distribution: (a) rainfall cumulated over 30-day moving windows; (b) mean daily background rainfall; (c) application to the 20-year reference period.

In the application phase, once h̅ has been determined using the procedure described above, it is distributed according to the three considered patterns: uniform, triangular, and parabolic. For each pattern, the value of P30 that, when combined with the corresponding predisposing rainfall profile, results in FS = 1 is identified. This value is then expressed as a percentile within the D1 distribution of daily rainfall observed in the study area. The pattern considered most suitable is the one requiring the minimum P30 value among those corresponding to a percentile higher than the 90th percentile, which is assumed as the reference threshold for triggering events. Figure 4 shows an example of a synthetic rainfall distribution and its comparison with the D1 distribution for the determination of the P 30 percentile.