unable to define variable Tobc_in

[xiaozhu557]

Hi everyone,
I had got an error message when creating a new file (*_avg.nc or *_rst.nc) while model running. The error message as follow:

Code: Select all

STEP   Day HH:MM:SS  KINETIC_ENRG   POTEN_ENRG    TOTAL_ENRG    NET_VOLUME
           C => (i,j,k)       Cu            Cv            Cw         Max Speed
 
1369440  4755 00:00:00  3.431246E-03  2.322748E+04  2.322748E+04  4.688233E+17
          (0208,244,22)  4.221437E-02  1.780488E-02  0.000000E+00  2.183069E+00
       DEF_HIS   - inquiring history file: np8_his_0013.nc
       DEF_AVG   - inquiring average file: np8_avg_0013.nc
 
  DEF_VAR - unable to define variable: Tobc_in
            in NetCDF file: np8_rst.nc
 ...........
All percentages are with respect to total time =         3186.795
  
 ROMS/TOMS - Output NetCDF summary for Grid 01:
 
 Analytical header files used:
    
     ROMS/Functionals/ana_btflux.h
     ROMS/Functionals/ana_hmixcoef.h
     ROMS/Functionals/ana_nudgcoef.h
 
  ROMS/TOMS - Output error ............ exit_flag:   3
 
 
 ERROR: Abnormal termination: NetCDF OUTPUT.
 REASON: NetCDF: Invalid dimension ID or name

I had run the model very well before this. I have not change any thing for it. So I think that would be the problem from the super-computer, but just by the error message the manager can not get any useful messages to check the linux system. So I am confused for that. Anyone know what's the possible reasons for that?

Thanks you for any suggestions.

[kate]

It looks like you are running from a restart? Is this your first restart? Somehow it is trying to define a variable in your restart file, which should only happen when this file has been newly created. What value do you have for NRREC?

[xiaozhu557]

It looks like you are running from a restart? Is this your first restart? Somehow it is trying to define a variable in your restart file, which should only happen when this file has been newly created. What value do you have for NRREC?

Thanks for your reply,Kate.

Yes, I am running from a restart with a history file as the initial file. And this is my first restart. But I have used this kind of setting to restart running in other model domain.

I have set some parameters in ocean.in file as follow:

! Time-Stepping parameters.
110
111 NTIMES == 93312000 ! 243d !14400 4m11d !10080 3m11d !275040 !269280 5y3m
112 DT == 300.0d0
113 NDTFAST == 10
114
115 ! Model iteration loops parameters.
116
117 ERstr = 1
118 ERend = 1
119 Nouter = 1
120 Ninner = 1
121 Nintervals = 1
122
123 ! Number of eigenvalues (NEV) and eigenvectors (NCV) to compute for the
124 ! Lanczos/Arnoldi problem in the Generalized Stability Theory (GST)
125 ! analysis. NCV must be greater than NEV (see documentation below).
126 !-------------- This is used for PROPAGATOR
127 NEV = 2 ! Number of eigenvalues
128 NCV = 10 ! Number of eigenvectors
129
130 ! Input/Output parameters.
131
132 NRREC == 72
133 LcycleRST == T
134 NRST == 288 !!288 !! 1day
135 NSTA == 0
136 NFLT == 0
137 NINFO == 1
138
139 ! Output history, average, diagnostic files parameters.
140
141 LDEFOUT == T
142 NHIS == 1440 !!5 days 72 !!6 hour
143 NDEFHIS == 105408 !! 1 year
144 NTSAVG == 1
145 NAVG == 288
146 NDEFAVG == 105408
147 NTSDIA == 0
148 NDIA == 2214
149 NDEFDIA == 2214

Please remind that I have got the same error message when it creating the average file, once I closed to output restart file with setting

NRST=0

Thanks for your help again.

[kate]

Code: Select all

132 NRREC == 72
133 LcycleRST == T
134 NRST == 288 !!288 !! 1day

A couple points about this:

* LcycleRST means to alternately write to records 1 and 2.
* NRREC == 0 is a special case for beginning a fresh start. I would extract the one record you want as your initial condition into a new file and set NRREC==0. Set ININAME to point to this file and let it use the np8_rst.nc name for restarts. It should create this anew, so save the old one somewhere else.
* The ROMS parsing of this file is tricky. It's not entirely safe to put stuff after a '!' because not all things after the '!' are ignored unless the '!' is at the start of a line.

[xiaozhu557]

Hi Kate,
I had restarted the model with a restart file np8_rst.nc many times, it worked well.
But I have to restart the model with a history file(np8_his_0013.nc) right now, because I have no the restart record in the restart file for that time point which I want. Now I have settings as follow:

Code: Select all

      GRDNAME == np8_grd.nc
 636      ININAME == np8_his_0013.nc
 637 !     ITLNAME == ocean_itl.nc
 638 !     IRPNAME == ocean_irp.nc
 639 !     IADNAME == ocean_iad.nc
 640 !     CLMNAME == np8_clm.nc
 641      BRYNAME == np8_bry_soda.nc
 642 !     FWDNAME == ocean_fwd.nc
 643 !     ADSNAME == ocean_ads.nc

I have uploaded the ncdump reutlt of the file np8_his_0013.nc, you can download and check it. Could you tell me whether I can use it to restart my model? If I can, how shoud I set the parameters in my ocean.in?
Thanks!