Abnormal Cooling at Shallow Water Area

[corvianawatie]

Dear All

I have a problem with my model. The bottom temperature is intruding to the shallow water area very rapidly (the bathymetry range is 60 - 5000m, with 30 vertical levels). For the longer simulation, the cooler temperature is spreading towards the surface. I activated the following CPP.

1. Previously I was thinking it is due to the lateral mixing. I already activate MIX_GEO_UV and MIX_GEO_TS, but the problems standstill. I tried using LMD_SAPHIRO, but the result still same.
2. Another problem might be due to the surface heatflux. But, the cooling only occured at shallow water area. The temperature profile at the open ocean is normal.
3. Do I need to nudging the bottom temperature to climatology?

I have seen the similar problem at another thread, but I cant find the exact solution. Please kindly help if you have any experience with this problem. Thanks.

Code: Select all

#define UV_ADV
#define UV_COR
#define UV_QDRAG
#define DJ_GRADPS
#define SPLINES_VDIFF
#define SPLINES_VVISC
#define TS_U3HADVECTION
#undef  TS_C4VADVECTION
#undef  NONLIN_EOS
#define SALINITY
#define SOLVE3D
#define MASKING
#undef  QCORRECTION
#undef  SRELAXATION
#define CURVGRID
#define AVERAGES
#define LMD_MIXING
#ifdef  LMD_MIXING
# define LMD_RIMIX
# define LMD_CONVEC
# define LMD_SKPP
# define LMD_NONLOCAL
# define RI_SPLINES
#endif
#define ANA_BSFLUX
#define ANA_BTFLUX

#define SPHERICAL
#define ANA_SMFLUX
#define MIX_GEO_UV
#define MIX_GEO_TS
#define TS_DIF2
#define UV_VIS2
#define IMPLICIT_NUDGING
#define RADIATION_2D
#define DIFF_GRID
#define VISC_GRID
#define LIMIT_VDIFF
#define LIMIT_VVISC
#define LIMIT_BSTRESS
#define BODYFORCE

Bottom Temp (Day-0)

Day-0 (initial)

d0.PNG (123.52 KiB) Viewed 7351 times

Bottom Temp (Day-30)

Day-30

d30.PNG (129.65 KiB) Viewed 7351 times

Bottom Temp (Day-60)

Day-60

D60.PNG (132.37 KiB) Viewed 7351 times

[kate]

Code: Select all

#undef  TS_C4VADVECTION
#define BODYFORCE

These and others of your choices put you into a realm where I have no experience at all. Have you done any bathymetry smoothing?

[corvianawatie]

I did bathymetry smoothing using B.C Saphiro constant available at GridBuilder.
Sorry Kate, I think I made a wrong copy. I used this CPP

#define UV_ADV
#define UV_COR
#define UV_QDRAG
#define DJ_GRADPS
#define SPLINES_VDIFF
#define SPLINES_VVISC
#define TS_U3HADVECTION
#define TS_C4VADVECTION
#define NONLIN_EOS
#define SALINITY
#define SOLVE3D
#define MASKING
#undef QCORRECTION
#undef SRELAXATION
#define CURVGRID
#define AVERAGES
#define LMD_MIXING
#ifdef LMD_MIXING
# define LMD_RIMIX
# define LMD_CONVEC
# define LMD_SKPP
# define LMD_NONLOCAL
# define RI_SPLINES
#endif
#define ANA_BSFLUX
#define ANA_BTFLUX

I modified from damee4.h and added this:

#define SPHERICAL
#define ANA_SMFLUX
#define MIX_GEO_UV
#define MIX_GEO_TS
#define TS_DIF2
#define UV_VIS2
#define IMPLICIT_NUDGING
#define RADIATION_2D
#define DIFF_GRID
#define VISC_GRID
#define LIMIT_VDIFF
#define LIMIT_VVISC
#define LIMIT_BSTRESS
#define BODYFORCE

[kate]

That's better, but still, what are your BODYFORCE parameters? I haven't used that option since DAMEE, which was a long time ago. It applies the surface stress not just at the surface, but over some number of points in the vertical. It was for a time when we couldn't resolve the surface boundary layer, with only 8 grid points in the vertical. I remember using it for a flat-bottomed domain where it makes more sense in terms of the problem you are seeing. Can you try turning it off?

[corvianawatie]

Ok Kate, I will try to turn off the BODYFORCE. I used surface heat flux and freshwater flux in the surface boundary layer.

[Kosa]

Where are you getting your net longwave heat flux? I've had issues with unrealistic temperatures in shallow regions when applying heat fluxes. The problem is that the heat flux depends on the temperature difference between the air and the ocean, but when we take heat fluxes from some other model (usually with a coarser resolution ocean which doesn't have the same land-sea mask as our ROMS domain), you can continue fluxing heat out of the ocean even though it's already colder than the atmosphere. Ideas: you could manually adjust the heat flux in your forcing file in the shallow regions (near land) or look into using bulk 8 forcing which will take into account the temperature of the ocean.

[corvianawatie]

Where are you getting your net longwave heat flux? I've had issues with unrealistic temperatures in shallow regions when applying heat fluxes. The problem is that the heat flux depends on the temperature difference between the air and the ocean, but when we take heat fluxes from some other model (usually with a coarser resolution ocean which doesn't have the same land-sea mask as our ROMS domain), you can continue fluxing heat out of the ocean even though it's already colder than the atmosphere. Ideas: you could manually adjust the heat flux in your forcing file in the shallow regions (near land) or look into using bulk 8 forcing which will take into account the temperature of the ocean.

My atmospheric forcing are obtained from ERA5 data, and the net heat flux is calculated using d_ecmwf2roms.m. As I know, the surface forcing will be interpolated into the ocean model grid. Ok, I will try your suggestion to using bulk fluxes.

[ymamoutos]

Greetings

Are you downloading ERA5 data from ECMWF site or
from Copernicus CDS (https://cds.climate.copernicus.eu)?

If your are downloading from 2nd source there is no need to use
d_ecmwf2roms.m script because the fields that need to be accumulated
are already accumulated according to ERA5 CDS manual. If you are using
ECMWF site then you must modify the provided script (d_ecmwf2roms.m)
to accumulate the radiation fields (shortwave, longwave, sensible),
evaporation, total precipitation and surface stresses for hourly and
not 3 hours timestep. If you are sure about how you create your forcing
files and still want to use the net fluxes try to use QCORRECTION option
using either an netcdf file or a fix value in an analytical file (ana_dqdsst.h).
But I think that the previous suggestion about using BULK_FLUXES option
is more safe.

GIannis