open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I have run a simulation of a Gulf region, where I have given tide as the boundary condition at the open boundaries.

Now I want to select a sub region of the Gulf and use the output from the global model as an input to the boundaries of the small region model.

What are the parameter which are derived from the Global model outputs to be imposed at the open boundary nodes of the small model?
The Global model outputs are U, V, S, B and H.

I understand that I have to modify the BORD subroutine where the values should be populated at the liquid boundaries.

Is there any example case illustration?

Regards,
Gourish

Hello,

This is a difficult task, maybe someone will give an example, even quick and dirty. If not you need to have the list of boundary points of the smaller mesh, with their coordinates, so that you can locate them in the larger mesh (subroutine derive in library bief will give you inspiration on how to locate a point in a mesh and find the barycentric coordinates). You will have then to rerun the larger model and write in a file the (interpolated) values of elevation and velocities, varying in time.

Then the smaller model can read this file and find what is to be prescribed at the boundaries, this probably must be done with Thompson boundary conditions to have a well posed problem.

The test case 028_ondem2 shows an example where the depth at boundaries is deduced from data in a file. It works in parallel and shows that you can nearly empty the standard BORD subroutine and write your own code, provided that the values of HBOR, UBOR and VBOR are given. In this case UBOR and VBOR are not given because the velocities are declared as unknown and the Thompson method will estimate them.

With best regards,

Jean-Michel Hervouet

Hi Jean-Michel,
Thank you for the reply.

There is an option in Blue-Kenue where the time series value can be extracted along a line or a set of points.

As you have mentioned that file has to be created giving the prescribe H, U and V values at the boundaries, therefore I have used the outer open boundary locations of the small region mesh as line, and extracted the appropriate time series value using BlueKenue in a file.

Using these extracted values from the file I will try the method as demonstrated by test case 028_ondem2.

Thanks again,
Gourish

Hi everyone,

As mentioned earlier I have extracted the tidal vales from the global model
at the open boundary location of the small sub region model.
The file are extracted in the file tide_at_bound.dat.
The values at the respective small domain boundary nodes are read using the BORD subroutine.

I am confused whether my implementation (on line no. 465, file ril.f) will work also in parallel or not. Secondly how can I impose the U and V values, Will it be proper to directly populated UBOR and VBOR values simultaneous along with HBOR after line 465 in ril.f?

I have attached the boundary condition file, the fortran file, the case file and the boundary values file.

Thanks,
Gourish

Dear Gourish,

I want to do exactly the same thing, i want to run my small model with boudary condition extract from the result of a large model.

I had tried to modify the bord.f but i do not know much about fortran programming.

Can you help me please.

REGARDS.
YASMINA

Hello Yasmina,

Thanks for asking.

Can you give a description of the model boundary forcing which you want to do in the model. For example like the information about the number and type of forcing. I normally force U, V, H derived from a large model into a small model at the boundaries. Based this I create a text file which will have the values of U, V & H extracted from the large model and to be forced in the small model. I use bluekenue to extract U, V & H at the boundary nodes of the small model from the large model.

Perhaps if you can post a dummy pictorial representation of the actual work which you want to do, then based on the information I can post my modified scripts for the latest version of Telemac for boundary forcing. You need not give the actual area and geometry representation, just a draft diagram should be fine. The fortran module which I have attached earlier is of the old version of Telemac, and thus I need to modify it as per the latest version.

Sorry I had been out of touch with Telemac for some months. However I can modify my old scripts and module and post it here which you can use it for your modeling case. Sometimes I need to modify the boundary data format extracted using Bluekenue using some python scripts, I shall post them along with the description as how to use them.

With Best Regards,
Gourish Salgaonkar

Hello Gourish,

Thank you for your reply.

Yes you right i will post what i had done before.

So actualy i have result of simulation with coupled model for a large model (The attached picture).

From this large model i want:
1- First to extract the velocity (UV) and free surface wich are varying in time and domain, to use them as a boundary condition for the small model.
2- Second i want to do the same thing but for result of TOMAWAC, so i want to use the hight significative wave varying in time and domain and use it at the boudary of the small model.


I already extracted the VELOCITY, the free surface and the hight significative waves (File attached "fichier de frontière") i make my own variable (ELV) with number of node corresponding with the small model between bracket of each point.

I tried to modifie the bord.f file but i am not very famillar with fortran langage.


I hope that all this information is sufficient and that you can help me to solve this problem.

Thank you in advance.
Best Regards,
YASMINA HAMDOUS.

Hello Yasmina,

I have understood your requirements. I had already done a similar simulation with Telemac earlier version. I may have to modify the bord.f subroutine as per the latest version.
The earlier files which I have posted are quite old, and I had coded a optimized subroutine for the boundary forcing. I shall post the program and Readme file which will help you in running the simulation. It may take some time, may be a day or two to get the module working, however I shall post it at the earliest. Perhaps I can also check if the latest version of Telemac has any option for supporting such problems.

I have not yet worked with TOMAWAC with the similar case. However I shall try to work on this problem and suggest you a possible solution.

Thanks, the information which you have provided is quite sufficient.

With Best Regards,
Gourish Salgaonkar

Hi,

Yes!!! the program and Readme file will help me so much, i will wait for your post.

In parallel i will try to learn more of Fortran Langage.

Realy thank you for your helps.

Best Regards,
YASMINA HAMDOUS.

Hi,

This is probably well over due, but any chance you can still send the modified BORD?


cheers,


NZ