open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi

Just to be fully sure, make a clean full compilation of the system, and try again to run the case.

Regards

Unbelievable, I tried to run with ifortmpi instead of gfmpi and now is running ...!
Hopefully gives good results...
Regards

Hi,

I still cannot understand where I have been wrong.
I saw that meteo.f changed a bit compared to v6.3 (MY_OPTION does not exist anymore, but is substitute by...?! )
Also I activated AIR PRESSURE in my cas file
I chose atmospheric exchange model option 1
I tried different ways to write the formatted file ...

but I always get the error:

===============================================
WARNING: TIME OF CALCULATION AT = 64800.0000000000
IS BIGGER THAN MAXIMUM TIME IN YOUR INPUT DATA
FILE T = 64800.0000000000
===============================================

PLANTE: PROGRAM STOPPED AFTER AN ERROR

---

Primary job terminated normally, but 1 process returned
a non-zero exit code.. Per user-direction, the job has been aborted.

---

---

mpiexec detected that one or more processes exited with non-zero status, thus causing
the job to be terminated. The first process to do so was:

Process name: [[51318,1],0]
Exit code: 2


which is actually not true... fo1 and computation last 1 day long (86400s)

And additionally I do not have mass at the time steps and through boundaries... I have NaN
It seems it is not reading correctly the input file

I attach cas, fortran and formatted file

Thanks for any suggestion

Best regards

Hi,

only to sum up. Maybe someone had a similar problem.

I am computing with version 7.1, using ATMOSPHERE EXCHANGE MODEL = 1. I have a formatted file with variable wind and temperature over 1 day (each 3 hours), I don't specify the resting variables, I only want to consider the variation of wind and temperature. I compute for one day saving each 3 hours.

I got results but no mass (NaN) each time steps after the first and at the boundaries (neither water nor tracer) - I attach one example of output file, which gives results, but wrong ones as I told





In the fortran file I have corrxy, bord3d, meteo, interpmeteo and condim (for refinement of layers near bottom and surface)

The cases with contant temperature of the air worked well and also the ones with variable wind.

I don't have ideas anymore to make work this case... Any help is highly appreciated!

Thanks in advance

Best regards

Hello,

In all such cases where you have a NaN, the best is to choose compiler directives that allow you to debug. For example with gfortran, I add:

-fbounds-check -ffpe-trap=invalid,zero,overflow -fbacktrace -Wall

and you will know where and at what line the first NaN occurs, generally a division by zero somewhere, then it will be much clearer to analyse the mistake.

This will require to recompile all the system if the mistake is not in your user Fortran. Be aware that such a configuration slows down the program, so it is not to be used always.

With best regards,

Jean-Michel Hervouet

Thank you, I will try asap and get back to the forum if I find out something interesting.
Nevertheless, I think the problem is that my fortran file is not reading entirely my text formatted file. I tried adding meteo.f and then it returns not anymore NaN, but it only calculates wind, and no temperature.
I still have to understand how to make telemac v7p0r1 working with ATMOSPHERE-WATER EXCHANGE MODEL=1 and giving only wind and temperature variable e.g. during a day.

With best regards