On the choice of the optimal frequency analysis of annual extreme rainfall by multifractal approach

Intensity Duration Frequency (IDF) curves are important tools for hydraulic and hydrologic design. Considering that there are different approaches to obtain the rainfall quantiles on which the IDF curves are based, the availability of a method to evaluate their reliability is of great importance. With this aim, in this work the multifractal properties of hourly rainfall data recorded at 23 rain gauges in the Umbria Region (Italy) have been used to select the most appropriate frequency analysis method of extreme annual rainfall at each location. Three methods have been tested: Local Frequency Analysis (by fitting at each station extreme annual rainfall data of different durations by a probability distribution function), Regional Frequency Analysis approach based on L-Moments and flood index method (considering the extreme annual rainfall data from all the stations), and a variant of the latter method known as In-site Regional Frequency Analysis (based on the consideration of the station as a region). Therefore, quantiles of rainfall for different durations and return periods have been obtained. These quantiles have been fitted to the Montana Intensity-Duration-Frequency (IDF) curve. The scaling properties of rainfall have been obtained through out the empirical moments scaling function K(q). Their comparison with some scaling behavior properties of the IDF curves has let the selection of the most adequate quantile estimation method at each site, being the Regional Frequency Analysis the most appropriate one for 14 out of the 23 sites included in the study.