Custom Query (964 matches)

The local variable Nfiles used to count the number of files per field(s) during record splitting was initialized before being allocated in inp_par.F. This caused unexpected behavior in some compilers that do not initialize variable to zero by default. Many thanks to Kate Hedstrom for bringing this to my attention.

I also added the following conditional when computing the Infinity variable that it is needed when coupling with SWAN:

#ifdef SWAN_COUPLING
      Infinity=one/zero
#endif

This is still not a solution but at least is not used everytime that we run ROMS in other applications. Many thanks to Mark Hadfield for ​reporting this.

If I run the model using a NRREC value different than 0, the subroutine def_his() always expects an existing NetCDF history and it results in:

DEF_HIS   - inquiring history file: ''/path/to/history.nc''

 NETCDF_OPEN - unable to open existing NetCDF file:
               ''/path/to/history.nc''
               call from:  def_his.F

The problem seems due to the missing initialization of the parameter idefHIS in Modules/mod_ncparam.F. Please, note that in releases prior to 553 this parameter was initialized to -1.

Looking at Nonlinear/output.F the following lines of code are never executed due to the limiting condition over idefHIS (which is never assigned, hence it's always 0):

    76            IF (idefHIS(ng).lt.0) THEN
    77              idefHIS(ng)=((ntstart(ng)-1)/ndefHIS(ng))*ndefHIS(ng)
    78              IF (idefHIS(ng).lt.iic(ng)-1) THEN
    79                idefHIS(ng)=idefHIS(ng)+ndefHIS(ng)
    80              END IF
    81            END IF

In the same subroutine the following if/then construct results in the assignment of NewFile which will be always .FALSE. whenever NRREC is not equal to 0:

    82            IF ((nrrec(ng).ne.0).and.(iic(ng).eq.ntstart(ng))) THEN
    83              IF ((iic(ng)-1).eq.idefHIS(ng)) THEN
    84                Ldefine=.FALSE.                 ! finished file, delay
    85              ELSE                              ! creation of next file
    86                Ldefine=.TRUE.
    87                NewFile=.FALSE.                 ! unfinished file, inquire
    88              END IF                            ! content for appending
    89              idefHIS(ng)=idefHIS(ng)+nHIS(ng)  ! restart offset
    90            ELSE IF ((iic(ng)-1).eq.idefHIS(ng)) THEN
    91              idefHIS(ng)=idefHIS(ng)+ndefHIS(ng)
    92              IF (nHIS(ng).ne.ndefHIS(ng).and.iic(ng).eq.ntstart(ng)) THEN
    93                idefHIS(ng)=idefHIS(ng)+nHIS(ng)  ! multiple record offset
    94              END IF
    95              Ldefine=.TRUE.
    96              NewFile=.TRUE.
    97            ELSE
    98              Ldefine=.FALSE.
    99            END IF

As a workaround I re-introduced the initialization of at least the following two parameters inside Modules/mod_ncparam.F:

      DO ng=1,Ngrids
        idefAVG(ng)=-1
        idefHIS(ng)=-1
      END DO

This lets the first aforementioned block of code (output.F: line 77) to correctly set the idefHIS(ng) parameter and in the subsequent if/then construct the NewFile parameter is permitted to be .TRUE.

Thank you, Alex

The following time-averaged CPP options are now available:

• AVERAGES: use if writing out time-averaged nonlinear model fields.
• AD_AVERAGES: use if writing out time-averaged adjoint model fields. Used primarily in adjoint sensitivity simulations.
• RP_AVERAGES: use if writing out time-averaged representer model fields. Used only in the R4D-Var algorithm (W4DVAR) to save the representer model time-averaged fields for each outer loop in separated files.
• TL_AVERAGES: use if writing out time-averaged tangent linear model fields. It will used in future to load averages of the TLM kernel.

Notice that only one of these CPP options can be activated at the same time. They are mutually exclusive! This is because the same internal arrays (declared in mod_averages.F) are uses to stored the accumulated field sums.

Now, the I4D-Var and 4D-PSAS algorithms (IS4DVAR and W4DPSAS, respectively) will write out the time-averaged and diagnostic fields for each outer loop in separated files if the AVERAGES and/or DIAGNOSTICS_TS/DIAGNOSTICS_UV options are activated.

Warning: The CPP directives associated with AVERAGES where modified in several routines (like biology and sediment) to include instead:

#if defined AVERAGES    || \
   (defined AD_AVERAGES && defined ADJOINT) || \
   (defined RP_AVERAGES && defined TL_IOMS) || \
   (defined TL_AVERAGES && defined TANGENT)

Many thanks to Andy Moore for suggesting this useful capabilities for the 4D-Var data assimilation algorithms. This will become handy in 4D-Var reanalysis of ocean circulation for a particular region.