Hydraulic modeling

[Jana Corominas]

Hello everyone,

I have been using wrf-hydro uncoupled from WRF (but using WRF as input data) for a while now with success to model streamflow. I would like to make a step forward into hydraulic modeling (to evalue for example flood risk). Has anyone used WRF-Hydro for such approach? Is there a way to use wrf-hydro output for this intention? For example as input data for other models (I am not familiar with existing hydraulic models).

Thank you!

Jana

[Ulysse Pasquier]

Hi Jana, 

This study might be of interest, they combine WRF-Hydro with HEC-RAS 2D:

https://www.mdpi.com/2073-4441/11/11/2328/htm

Best,

Ulysse

[Javier Alvarez]

Thanks both, this is a very interesting paper and research field. 

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/1336fb6d-6706-4d22-b8f7-91dd732b6183n%40ucar.edu.

[Mengye Chen]

Hi Jana,

     I have fed the runoff from WRF-hydro to ANUGA model for 2D simulation, which is essentially Noah-MP coupling with ANUGA. I do not recommend the 1D to 2D scheme, since your only choice is diffusive wave routing and that doesn't perform so well. 

Mengye Chen

To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/CABLVm6Owbw-ij4zmb2DN84---tzXM2MV%2Bp9oTFGutx-fqt2VpA%40mail.gmail.com.

--

Mengye(V) Chen

 

[Jana Corominas]

Hello Mengye Chen,

I have looked for ANGUA information. I see it is a python based code but I found it quite difficult to model the domain and set up scenario with the data used in wrf or wrf-hydro model. For example if a bathymetry map is needed I wouldn't know where to start of how to work with it in a notebook.

I would like to use wrf + wrf-hydro input/ouput for the hydraulic modeling, is it really possible with ANUGA or does it require an extensive previous pre-processing of data to make it work? 

Jana

[Mengye Chen]

Hi Jana,

     You are coupling WRF-hydro output with another hydraulic model, of course, you need to prepare all the input for ANUGA all over again. You need a much finer spatial resolution and temporal resolution of your input data to make sure the explicit 2D shallow water equation would be stable.

     And yes, you need to use a higher resolution DEM and create triangular mesh on top of it. You need to convert the WRF hydros runoff (which is technically Noah-MP runoff) into ANUGA accepting format, which they have code for that. ANUGA support point flow input, if you want to go that route, then you just need the streamflow at one point in csv file as your forcing input.

    Wanna something that people couples land surface model and 2D water routing for you already? Try CREST-iMAP or LISFLOOD-FP. 

Mengye Chen

--

Mengye(V) Chen

 

[Jana Corominas]

Hello again Mengye Chen, thanks for your good explanation. I will investigate ANUGA model. 

One last question, is it possible to automate ANUGA model? I am running WRF-Hydro twice a day in a virtual machine using a bash script, and I would like to add an hydraulic model afterwards without human intervention.

Thanks again, 

Jana.

[Mengye Chen]

Hi Jana,

     Yes, ANUGA is a series of python functions. You can totally use python to call your bash script, followed by the ANUGA data preparation and ANUGA simulation functions. As long as your domain doesn't change, this should be easy. Make sure you have a functional hard drive that allows your Temp folder to I/O some large data between WRF hydro and ANUGA.

     It can be done even with a dynamic domain, but I have not done that yet.

     Depending on your temporal resolution and domain size of the 2D simulation, the hydraulic part might take many hours. I definitely don't recommend you simulate a large continent. A small basin should be ok.

Mengye Chen

Center for Analysis and Prediction of Storms

University of Oklahoma

--

Mengye(V) Chen

 

[Jana Corominas]

Hello Mengye Chen, 

I'm working with a basin in Panama with an extension of 1300km2. The domain in WRF-Hydro is initially a 1km x 1km horizontal resolution grid with a finner mesh of 250m x 250m to do the routing. Do you think it will be very expensive computationally?

[Mengye Chen]

Hi Jana,

     Yes, 1300km2 should be fine.

     The 2D hydraulic simulation requires much finer DEM. 10m or lower would be better. I'm not sure if WPS would be able to provide you with such spatial resolution. But I think you can try.

--

Mengye(V) Chen

 

[Kevin]

WPS should not have difficulty creating a grid of this size and resolution. You will of course need to generate the WPS tiles for an elevation dataset of that resolution, however. 

Kevin

[Jana Corominas]

Hello again, I managed to get a 30m resolution DEM for my domain of interest. 
After reading the user manual and different example case studies, I'm not sure how to force the model to start with wrf-hydro output runoff as initial condition.

Is "stage" the variable to be specified in this case? Is it water height (m) above ground? 
Is there a way to give streamflow (Q, m3/s) in all the domain and force it to the model? 

Mengye Chen used Wrf-hydro runoff for this purpose but I don't know how to start...

Thanks in advance,

Jana.

[Mengye Chen]

Hi Jana,

     If you are trying to use the stream flow as the forcing, please take a look at this sample code I made when I was using ANUGA for testing.

     "Stage" is the water surface from the sea level. You need to subtract the stage by elevation to get the depth of water on the land surface.

Mengye Chen

--

Mengye(V) Chen