Very cold water forming at boundaries

[JDTilley]

I am still a novice at ocean modeling, and I am having an issue with cold water (< 5º in shallow water) forming at my boundaries during August in the Gulf of Mexico. I have checked and double checked my climatology, boundary, and forcing files, but can't see anything that should cause this. I have my boundary conditions sets as:

Code: Select all

   LBC(isFsur) ==   Che     Che     Che     Che         ! free-surface
   LBC(isUbar) ==   Shc     Shc     Shc     Shc         ! 2D U-momentum
   LBC(isVbar) ==   Shc     Shc     Shc     Shc         ! 2D V-momentum
   LBC(isUvel) ==   Rad     Rad     Rad     Rad         ! 3D U-momentum
   LBC(isVvel) ==   Rad     Rad     Rad     Rad         ! 3D V-momentum
   LBC(isMtke) ==   Per     Per     Per     Per         ! mixing TKE

   LBC(isTvar) ==   RadNud  RadNud  RadNud  RadNud \       ! temperature
                    RadNud  RadNud  RadNud  RadNud         ! salinity

The problem begins right at the beginning of the model run. Does anyone know what might cause this? Any help would be greatly appreciated. I would attach some short mp4 animations, but it doesn't seem that I am allowed.

[jcwarner]

is the cold water forming at the boundaries of the model grid, or the land/sea boundary?

what are the Tnudg values at those locations? The climatology temp values may be = 0 because you have a finer grid than the clim data (just a guess).

[JDTilley]

The cold water is forming at the ocean boundaries, not at the land/sea boundaries. I have TNUDG set to 2*0.0d0, which now that I look at it, seems to imply zero. Could that be my problem? As far as my grid goes, it is the same resolution as my climatology and boundary files, although finer than my forcing file.

[JDTilley]

So I figured out you can upload animated gifs, so I am attaching a couple here. I am only doing short model runs at this point. It is the bottom layer that the cold is creeping in, and the surface has the opposite problem. I did try setting my nudging timescales, but the problem still remains:

Code: Select all

! Nudging/relaxation time scales, inverse scales will be computed
! internally, [1:Ngrids].

       TNUDG == 2*10.0d0                    ! days
       ZNUDG == 0.0d0                      ! days
      M2NUDG == 0.0d0                      ! days
      M3NUDG == 10.0d0                      ! days

! Factor between passive (outflow) and active (inflow) open boundary
! conditions, [1:Ngrids]. If OBCFAC > 1, nudging on inflow is stronger
! than on outflow (recommended).

      OBCFAC == 1.0d0                      ! nondimensional

I'm hoping this equates to an inflow timescale of 1 day and outflow timescale of 10 days, as in Fennel et al. (2009), but I might be wrong.

[kate]

I'm hoping this equates to an inflow timescale of 1 day and outflow timescale of 10 days,

Actually, you have them both set to 10 days. How are you setting the boundary values?

[JDTilley]

My boundaries were derived from my climatology file using SeaPy. I created an animation of the cross-section of the southern boundary and everything looked reasonable. So how would I set my inflow/outflow to 1 and 10 days? Thanks.

[kate]

Just set OBCFAC to ten to get what you want. It's the ratio of the outflow and inflow time scales.

[JDTilley]

Thanks. I set my nudging coefficients as:

Code: Select all

! Nudging/relaxation time scales, inverse scales will be computed
! internally, [1:Ngrids].

       TNUDG == 2*10.0d0                    ! days
       ZNUDG == 0.0d0                      ! days
      M2NUDG == 0.0d0                      ! days
      M3NUDG == 0.0d0                      ! days

! Factor between passive (outflow) and active (inflow) open boundary
! conditions, [1:Ngrids]. If OBCFAC > 1, nudging on inflow is stronger
! than on outflow (recommended).

      OBCFAC == 10.0d0                      ! nondimensional

I also realized I didn't have

Code: Select all

#define TCLM_NUDGING

set in my header file. However, I am still having problems. Any ideas how I can fix my problem?

[JDTilley]

Actually, looking over my ocean.in file it seems that the TCLM_NUDGING macro (among others) is obsolete. So I will try to turn on the logical switches for LTracerCLM and LnudgeTCLM and defining ANA_NUDGCOEF in the header file.

EDIT: scratch that, now it gives an error:

Code: Select all

 NETCDF_GET_FVAR_1D - error while inquiring ID for variable:  temp_time
                      in input file:  my_clim_ms_sound.nc
                      call from:  get_cycle.F

[JDTilley]

Well the error above was cause by me not having set temp:time in my climatology file. So I managed to get the model running again, but I still have the same problem. Any insights would be greatly appreciated.

[JDTilley]

I think I've figured out my problem. I think the data arrays in my boundary files are backward. At first I thought the lines at the edges of my animations were artifacts of my laziness. But that is the actually output in my _his file. It seems the script I used to create my boundary file was expecting the climatology data in the inverse order. I guess there's only one way to find out at this point. At least I learned a lot about nudging