open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I'm realizing the simulation of the river "Gave de Pau" in Orthez (France) and my results are really coherent with results from other studies and with flood traces (from 2013) for high flow.

But when I run the simulation in low flow, the results are very far from the reality (over 1 meter above the reality).

Do I need to change the parameters to run the simulation for low flow?
I attach my case folders so that you can have a look. I aldready made a difference in the parameter file by removing the Thompson condition when I run the simulation for low flow.

Thank you for your help.

Hi

case folder is missing!
think to zip the files to avoid problems in upload

regards

Hi,

Sorry the files were too big even zipped. Here are the parameters files, I hope it works.

Hi
It's difficult to try to explain where the problem could comes from with only the steering file...
We don't need the full result file, the last time step if you're running in steady flow or the maximum could be sufficient.
Otherwise, the geometry and the BC file could allow us to run the case.
Usually, between high and low flow, the friction or turbulence parameter could change a little but steering file are more or less the same...

Thank you for your answer. Here are the files to run the simulation.

I think the problem could come from my rating curve which is not precise enough for low flow.

Hi
About the rating curve, this could play a role but in your case it will leads to a downstream water level around 45.3 so I don't know if it's right or wrong...
Could you give us an idea of the "right" water level you are supposed to find?
Regards

The "right" water level is 49,5 at the downstream boundary for a flow of 88,7 m3/s.

That's why I'm saying that my rating curve is false, on a first approach for flood I used a rating curve that was not optimized for low flow and I didn't think to change on a second approach.

Hello,

It would be interesting to see what is the equivalent size of roughness for your friction at low flow. In some cases, for example estuaries, the ripples that create the friction are not the same at different hours because they depend on the flow condition. This would then require a friction varying with time.

With best regards,

Jean-Michel Hervouet

The rating curve is really false in that case!
49,5 --> 472 m3/s

What I don't understand here is how you could find water level 1m above the reality with a such low level downstream...
Is it necessary for me to run the case or would you try to change your data and test?

Yes, it's really false. I was only interested in high flows at first and I didn't pay attention.

I don't understand either how the water was only 1 m above the reality, the rating curve was not respected.

Here is the folder with the right rating curve. If you don't mind, you can run the case to see if you find an answer.