[MITgcm-support] notUsingX/YPeriodicity

[Pochini, Enrico]

Thank you Martin and Estanislao,

in my setup I only have a full Northern OB, Southern closed boundary and
West and East partial land, partial sea.
My problem was that I saw values from the W boundary repeated on the
Eastern boundary for eta, which is not possible to specify as OBetaFile
with my options. So I was wondering if there was this periodicity at work.
Moreover I experienced an error at initialization of OBCS, a mismatch
between the OB which I specified and the mask, where I think the East OB
was checked against the land/sea mask on the West side. Setting either
insideOBmaskFile, or specify also the OB files on West side seem to have
solved the issue.

Kind regards,

Enrico

Il giorno mer 13 nov 2019 alle ore 10:25 Martin Losch <Martin.Losch at awi.de>
ha scritto:

> Hi Enrico,
>
> I am not sure what you are trying to do. The “standard” way of supressing
> the periodicity is to impose one wall at either end of the domain to arrive
> at a channel or, if you do it for both horizontal directions, a box with
> solid walls.
>
> If you want to use open boundaries (i.e. the pkg/obcs), then the
> periodicity is still there as the default behaviour, but it is overwritten
> by the obcs values. The “halos” may still contain values based on the
> periodicity, but they are never used.
>
> Martin
>
> > On 13. Nov 2019, at 07:34, ESTANISLAO GAVILAN PASCUAL-AHU <
> e.gavilan at hhu.edu.cn> wrote:
> >
> >
> > Hi Pochini,
> >
> > I am no sure about that option, but if you do not want to have a
> periodic open boundary (i.e. time dependent). You could just set the
> forcing period to zero. From the data file you could write this:
> > periodicExternalForcing= .FALSE.,
> > externForcingPeriod= 0.,
> > externForcingCycle = 120000.,
> > or from the the data.exf:
> > &EXF_NML_OBCS
> >
> > obcsNstartdate1=19781216,
> > obcsNstartdate2=180000,
> > obcsNperiod=0,
> > obcsNrepCycle=120000,
> > obcsSstartdate1=19781216,
> > obcsSstartdate2=180000,
> > obcsSperiod=0,
> > obcsSrepCycle=120000,
> >
> > Kind regards,
> >
> > Estanislao
> >
> >> Hi all,
> >>
> >> I am trying to deactivate the default behaviour of periodic boundaries.
> I
> >> tried setting in eedata notUsingXPeriodicity=.TRUE.,
> >> notUsingYPeriodicity=.TRUE., but the option is not recognized. Setting
> >> forcefully in eeparms_set.F, where it appears in COMMON block EEPARMS,
> >> cause the following error to appear:
> >>
> >> *** An error occurred in MPI_Cart_rank
> >>> *** reported by process [3771400193,0]
> >>> *** on communicator MPI_COMMUNICATOR 3
> >>> *** MPI_ERR_ARG: invalid argument of some other kind
> >>> *** MPI_ERRORS_ARE_FATAL (processes in this communicator will now
> abort,
> >>> ***    and potentially your MPI job)
> >>> 6 more processes have sent help message help-mpi-errors.txt /
> >>> mpi_errors_are_fatal
> >>> Set MCA parameter "orte_base_help_aggregate" to 0 to see all help /
> error
> >>> messages
> >>>
> >>
> >> The two lines
> >> #define CAN_PREVENT_X_PERIODICITY
> >> #define CAN_PREVENT_Y_PERIODICITY
> >> are defined in CPP_EEOPTIONS.h
> >>
> >> Does the periodic domain definition interact with OB specifications?
> >> Is there a way to solve it without setting bathymetry=0 at margins?
> >>
> >> Thanks,
> >>
> >> Enrico
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20191113/af169b87/attachment.html>