Custom Query (986 matches)

The routine wpoints.F was updated to use the pmask_full, rmask_full, umask_full and vmask_full instead of the regular mask arrays (pmask, rmask, umask, and vmask). The _full arrays are used to ensure that masks at mass point source locations (river runoff) are set to water to avoid masking with _FillValue at those locations in output NetCDF files.

Recall that the WRITE_WATER option is used in ROMS to write output fields as a long vector (compact 1D array). The masked land points are not written to compress the size of output NetCDF files. I have not used this option in several years because it is not CF compliant and such ROMS output NetCDF cannot post-processed with third-party software.

Recall that the WRITE_WATER option cannot be used in conjuntion with wetting and drying WET_DRY because the 1D arrays witl change in size and violates the defined compact dimensions in the output NetCDF files.

Many thanks to David Ullman for ​bringing this to my attention.

There is a bug in uv_rotate.F when ASSUMED_SHAPE is activated. We need to have:

      real(r8), intent(inout) :: Uout(LBi:UBi,LBj:UBj,LBk:UBk)
      real(r8), intent(inout) :: Vout(LBi:UBi,LBj:UBj,LBk:UBk)

instead of

      real(r8), intent(inout) :: Uout(LBi:UBi,LBj:UBj,LBk:UBk)
      real(r8), intent(inout) :: Vout(LBi:UBi,LBj:UBj,LBk,UBk)

There were couple of issues with the output observations NetCDF file (MODNAME) used in 4D-Var and DART (Kalman filter) to report the model values at the observation locations and other important 4D-Var variables. We need to correct the survey dimension since it was one value too many. And correct the number observations with the same survey time (Nobs). This variable has the wrong dimension (state_var) with the wrong values. It was confused with new variable Nused_obs containing the number of usable (viable) observation per each state variable.

dimensions:
        record = 2 ;
        survey = 13 ;
        state_var = 7 ;
        cost_var = 8 ;
        datum = 6815 ;
variables:
        int Nobs(survey) ;
                Nobs:long_name = "number of observations with the same survey time" ;
        int Nused_obs(state_var) ;
                Nused_obs:long_name = "Number of usable observations" ;

Many thanks to Andy Moore for bringing this to my attention.