Various properties of the annual maximum rainfall depths, Hd, have to be clarified to meet specific objectives of practical interest. In this paper three different issues have been considered, assuming the Umbria region (central Italy) as case study. 1. Definition of the statistical distribution that adequately represents the samples of the annual maximum rainfall depths with duration d at regional scale. The analysis involved common statistical distribution types: Log-Normal, Gumbel and GEV. Samples composed of the Hd values observed in an entire region are fairly well represented by the GEV distribution. Samples relating to very short durations (5-10 minutes) are represented with significant difficulty. 2. Analysis of the hyetograph shape for events representing annual maximum rainfall depths. Knowledge of the most likely shapes of the Hd hyetographs is very important during the design rainfall determination as well as can improve the phase of "correcting" the underestimate error due to coarse temporal aggregation of rainfall data (Morbidelli et al., 2017). Very short intense events (with duration equal to 5-10 minutes) are frequently characterized by rectangular shapes, while as the duration increases, triangular and undefined shapes become more frequent. 3. Spatio-temporal analysis, on a regional basis and for all durations of practical interest (5 minutes-48 hours), of the 100 larger Hd elements. Knowing their temporal evolution allows to establish whether climate change is affecting them; knowing their spatial position allows to check if some areas are most affected by these extreme events. The temporal analysis, after the normalization with the number of working rain gauges, shows that there are no clear effects of climate change on the frequency of these events. On the other hand, considering only the last 40-50 years, an increasing trend of frequency could be deduced. Finally, the spatial analysis conducted over the study region shows that very short extreme events (with duration in the interval 5-15 minutes) appear to be more spatially uniform than events with long durations.
Characterization of Annual Maximum Rainfall Depths at Regional Scale
Renato Morbidelli
;Corrado Corradini;Jacopo Dari;Alessia Flammini;Carla Saltalippi
2020
Abstract
Various properties of the annual maximum rainfall depths, Hd, have to be clarified to meet specific objectives of practical interest. In this paper three different issues have been considered, assuming the Umbria region (central Italy) as case study. 1. Definition of the statistical distribution that adequately represents the samples of the annual maximum rainfall depths with duration d at regional scale. The analysis involved common statistical distribution types: Log-Normal, Gumbel and GEV. Samples composed of the Hd values observed in an entire region are fairly well represented by the GEV distribution. Samples relating to very short durations (5-10 minutes) are represented with significant difficulty. 2. Analysis of the hyetograph shape for events representing annual maximum rainfall depths. Knowledge of the most likely shapes of the Hd hyetographs is very important during the design rainfall determination as well as can improve the phase of "correcting" the underestimate error due to coarse temporal aggregation of rainfall data (Morbidelli et al., 2017). Very short intense events (with duration equal to 5-10 minutes) are frequently characterized by rectangular shapes, while as the duration increases, triangular and undefined shapes become more frequent. 3. Spatio-temporal analysis, on a regional basis and for all durations of practical interest (5 minutes-48 hours), of the 100 larger Hd elements. Knowing their temporal evolution allows to establish whether climate change is affecting them; knowing their spatial position allows to check if some areas are most affected by these extreme events. The temporal analysis, after the normalization with the number of working rain gauges, shows that there are no clear effects of climate change on the frequency of these events. On the other hand, considering only the last 40-50 years, an increasing trend of frequency could be deduced. Finally, the spatial analysis conducted over the study region shows that very short extreme events (with duration in the interval 5-15 minutes) appear to be more spatially uniform than events with long durations.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
