George,  

I think you can find quite comprehensive answer to your question in the presentation of David Gochis from NCAR, about the example of the Genoa flash flood.

You can full presentation under this link:

https://ams.confex.com/ams/93Annual/webprogram/Paper219864.html

As far as i know, WRF-HYDRO formally support Noah land surface model only right now. So, I guess that it's applicability for flood is the same as Noah model.

As long as WRF is being used to provide forecasts, I don't see why WRF-Hydro couldn't be used ot forecast flood events. Of course, based on atmospheric predictions made by WRF.

Which variable is appropriate to provide flood forecasts? Stream flow or head? Nevertheless, in a resolution of 100m which is initially recommended? I think that the investigation of the location of the flood is difficult.

George,  

I think you can find quite comprehensive answer to your question in the presentation of David Gochis from NCAR, about the example of the Genoa flash flood.

You can full presentation under this link:

https://ams.confex.com/ams/93Annual/webprogram/Paper219864.html

WRF-Hydro is in initial stage and now can be coupled with Noah and CLM as well. You ask about the flood forecasting using WRF-Hydro, which is very general question. I think WRF-Hydro could be more useful in predicting flash floods as compared to riverine floods. Similarly, If the upstream has complex snow melting mechanism then we may need to consider it's dynamics for better forecasting which wrf-hydro does not support. Furthermore, If you would like to use it for operational purposes then I must say, It would not be a wise approach but of course you can use it for research purposes. 

Yes, it can be used as flood forecasting model as it provides integrated modular system for hydrological processes. Here you can find the first publication about WRF Hydro, which was applied for flood forecasting purposes.   

Thank you very much Dr. Yucel. I have already read your publication and i found it very interesting. It is very useful mainly for calibration and flood investigation purposes. As though, the feasibility of many calibration ways and the variety of available forcing data for WRF-Hydro, give the opportunity to the users to make integrated flood and water management hydrological systems. Is there any way to investigate the flooded areas? For instance, if the stream flow exceeds the mean values, a flood is possible at that area. Nevertheless, where the runoff is spread? Can be the maximum river depth included as an input variable in the estimations?

Thanks for reading the article. Geo-morphology of the basin and river geometry which varies with location affect the river flooding. You can control the river geometry and its water depth (initial) in parameter file and manning parameter is critical parameter determines the velocity and depth of river. There is no direct input variable for maximum river depth.   

WRF-Hydro is now being used in the operational National Water Model of the United States for river forecasts of up to 10 days. More information can be found at http://water.noaa.gov/about/nwm . Over the last two years there were many extensions made to the code including support of Noah-MP (which gives better results for snow). Overbank flow is still an unsolved issue in the model but enhancements are planned for the future.

http://water.noaa.gov/about/nwm

Since Spring 2016, there is an interactive Google Map based tool for 5-day WRF-Hydro forecasts. It is configured for 2 Greek rivers. The project also includes weather forecasts and it was implemented by HCMR. http://meteo.hcmr.gr/hcmrweather/interactive_map.php

http://imbriw.hcmr.gr/en/tools/

This is nice work. Are there also results from Hydro part like surface runoff, streamflow?

Thank you very much Dr. Yucel. The discharge estimations are mainly based on the streamflow outputs. The model produces many variables like surface runoff, river depth (head), soil parameters etc. The exploitation of the additional parameters is a prospective plan. Do you know when the WRF-Hydro 4.0 will be released and its additional advantages?

As Dave told me in a few months a new version will be released with some updates (additions) in reservoir routing. This is important if the basin contains dams/reservoirs.

I wonder if WRF-Hydro experienced persons could share their experience on the possibility of numerical instability (as well as time step constraints and computational load) in the hydrologic model of the WRF-Hydro and ways to overcome that during the real time simulations. Along the same line, experiences with flood forecast lead time as one the most important criteria in flood warning.

Nice work. In my opinion this is good software and can be used for forecasting of flash floods.

I believe that WRF Hydro May be used for flow and flood prediction given the right setup. We have been using WRF rainfall integrated with hydrological models to predict floods at river basin scale and we have achieved satisfactory performance. With the additional setup of bias and residue corrections, the flow forecasts are pretty accurate