Runoff calculation for an extreme flood event in central Israel using different precipitation sources

1. Introduction Flash floods in semi-arid areas in the Eastern Mediterranean region can cause severe damages to property and leave behind casualties. This study presents a case study of a severe flood event that took place in central Israel during the week of 04/01/13 to 11/01/13, with peak discharges measured on the 08/01/13. Heavy rainfall up to 300mm, produced by a deep Cyprus low synoptic system, caused floods with a return period of 100-200 years (probability of 0.5% to 1%) in some of the basins (medium size basins) in central Israel. This event was unusual in terms of precipitation amounts, precipitation duration (over 6 days of continuous precipitation), intensities and precipitation volume. The Israeli Hydrological Service, who is responsible for the runoff monitoring system in the country, operated the WRF-HYDRO model (NDHMS-NCAR distributed Hydrological Model System) for flood forecasting. 2. Methodology The WRF-Hydro has a coupling architecture designed to simplify the coupling of terrestrial hydrological models with the WRF model (Weather Research Forecast). The NDHMS was developed to represent the spatial redistribution of overland flow and shallow saturated subsurface soil moisture. It is designed to simulate land surface hydrology and energy states and fluxes at a fairly high spatial resolution (typically 1 km or less) using physics based approaches. Overland flow is calculated when the depth of water on a model grid cell exceeds a specified retention depth. The WRF model domain was set for 300X300 km with a spatial resolution of 3 km (see figure 1) based on the GFS for initial conditions. 3. Preliminary results - Simulated stream flow vs. the observed The 7 days flood event (04/01/13 to 11/01/13) was simulated using the WRF-NDHMS model. The model produced hourly discharge for the full Israeli domain. The results were compared to the observations from the Israeli Hydrological Service (IHS) hydrometric stations network. Figure 4 presents the simulated hydrograph for "Amud" River, a relatively small channel in the Eastern slopes drainage to the lake of Galilee (see the basin location in figure 3). It can be seen in the figure that the simulated peak discharge in this event was very similar to the observed (75 m/sec2). This discharge is the highest that was ever measured in this stream (since 1940) and according to the IHS reports its return period is 1:200 years. However, it can also be seen that the timing of the simulation is of limited success: the raising of the NDHMS simulated peak discharge was 6 hours earlier than the actual peak captured by the hydrograph. The operation of the NDHMS model will be enhanced with observed precipitation data and with precipitation data derived from two radars (a regular and a polarimetric radar) operated for the Israeli Cloud Seeding Program by the Rain Enhancement Department of EMS Mekorot Projects, a subsidiary of Mekorot, the Israeli national water company. The polarimetric radar was recently installed and commissioned in northern Israel, on top of a 900 meters high mountain that commands all the major watersheds in the north. The 350KW systems is configured for rapid scanning of the atmosphere in a 165Km radius, and its rainfall estimation algorithms are using dual-pol moments to produce more accurate results in comparison to single-pol radars. The results from those 3 runs (NDHMS derived by NWP, radar and observed precipitation) for different basins in the domain will enable the examination of the advantages of running a high resolution hydrological model derived from a numerical weather prediction model, vs. simulations and predictions using radar and observed precipitation from automatic rain gauges.