1. Raganello Creek (Calabria, Italy) flash flood on August 20, 2018: Preliminary results.
- Author
-
Furnari, Luca, Avolio, Elenio, Senatore, Alfonso, Calidonna, Claudia R., and Mendicino, Giuseppe
- Subjects
- *
DOPPLER radar , *RIVERS , *RAINSTORMS , *RUNOFF , *MOUNTAINS , *RAINFALL , *FLOODS - Abstract
Due to its particular position in the middle of the Mediterranean Sea and its complex andsteep orography, the Calabrian peninsula (southern Italy) is frequently affected by severeprecipitation events and is particularly prone to significant ground effects. In the earlyafternoon of August 20, 2018, a storm cell caused a flash flood that affected the smallRaganello Creek, in the northern side of the region, causing 10 casualties at the outlet of theRaganello Gorges, where the extension of the catchment is about 100 km2. Though theregional monitoring network is rather dense, in the surroundings of the disaster site raingauges were not able to measure the very intense and highly localized rainfall. However, thethree-hour Surface Rainfall Total (SRT) provided by a single polarization Doppler radar ofthe Italian National Radar Network showed a storm cell with rain peaks of theorder of 70-100 mm/hour in a small mountain area not covered by the monitoringnetwork. The objective of this work is to evaluate the ability of the state-of-the-artmeteorological/hydrological modelling in simulating the Raganello Creek event using bothone-way and fully-coupled approaches, by means of the WRF/WRF-Hydro modellingsystem. The Advanced Research WRF (ARW) Model, version 3.9.1 is used with two one-waynested domains at the resolution of 10 km and 2 km respectively, while the resolution ofthe hydrological model, based on WRF-Hydro version 5.0, is 200 m. The initialand boundary conditions are provided by the ECMWF’s Integrated ForecastingSystem (IFS), in its deterministic forecast version at 9 km resolution. Several WRFsimulations have been carried out, before choosing the best configuration, to testthe impact of the initial conditions, horizontal resolutions and parameterizationschemes. Results show that the model, particularly in its fully-coupled version, is able to predictreasonably well the occurrence of the event both in terms of rainfall, thus identifying thestorm cell, and in terms of hydrological impact, reproducing timely what mighthave been (unfortunately, observed runoff is not available) the event hydrograph. [ABSTRACT FROM AUTHOR]
- Published
- 2019