Regional Frequency Analysis (RFA) is often used in hydrology to obtain rainfall estimates in places short data series. In this work RFA has been applied to extreme rainfall data series (with durations from 1 to 24 hours), with 23 rain gauges at the Umbria Region (Central Italy). The most important step in RFA is the creation of homogeneous groups of stations. The multifractal behaviour of hourly rainfall has resulted to be a very useful characteristic when joining stations into regions. Thus, a multifractal analysis of the hourly rainfall data series available for the 23 stations (from 1992 to 2013) has been carried out and the empirical moments scaling exponent functions K(q) have been obtained. Two parameters from the K(q) functions have been the bases for grouping the stations into regions when necessary. From the heterogeneity test of RFA, it has been checked that all the 23 stations form a homogeneous region for all the durations except than for 6 hours. A cluster analysis by means of multifractal parameters has then been performed for this duration. Two homogeneous subregions have been obtained with only one site excluded, being the local frequency analysis the one to be used for this place.
Regional Frequency Analysis of extreme rainfall in the Umbria Region (Central Italy)
CIFRODELLI, MARCO;MORBIDELLI, Renato;SALTALIPPI, Carla;
2016
Abstract
Regional Frequency Analysis (RFA) is often used in hydrology to obtain rainfall estimates in places short data series. In this work RFA has been applied to extreme rainfall data series (with durations from 1 to 24 hours), with 23 rain gauges at the Umbria Region (Central Italy). The most important step in RFA is the creation of homogeneous groups of stations. The multifractal behaviour of hourly rainfall has resulted to be a very useful characteristic when joining stations into regions. Thus, a multifractal analysis of the hourly rainfall data series available for the 23 stations (from 1992 to 2013) has been carried out and the empirical moments scaling exponent functions K(q) have been obtained. Two parameters from the K(q) functions have been the bases for grouping the stations into regions when necessary. From the heterogeneity test of RFA, it has been checked that all the 23 stations form a homogeneous region for all the durations except than for 6 hours. A cluster analysis by means of multifractal parameters has then been performed for this duration. Two homogeneous subregions have been obtained with only one site excluded, being the local frequency analysis the one to be used for this place.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
