open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

It seems that when I try to use unsteady user-defined wind fields (STATIONARY WIND=FALSE and WINDS FILE FORMAT=4), the variable NVWIN (number of wind variables to be read) is not defined during the initial stage of Tomawac (wac-->condiw-->lecdoi). The subroutine lecdoi.f does not assign any value to NVWIN (NVAR) in the case of INDIC=4.
With Tomawac alone, the model runs normally, but I'm afraid that print statements of NVWIN differs in function of the different places they are written in the code... (3,4,8...)
With a Tomawac-Telemac-Sisyphe configuration, I clearly have a segmentation fault during time loop of tomawac in wac-->semimp-->noudon (loop over NVAR at the end of the subroutine)

Can somebody help to find a patch (maybe in the initial call to lecdoi), so as to assign the right value to NVWIN ?

Sincerely,
Laurent Leballeur

It seems that a hard-coded value of NVWIN=2 in wac.F solves the problem in this case.

Regards,
Laurent Leballeur