Abnormal temperature along the boundary

[txu68]

Hi,

I am running a tidal simulation with river discharge. On the day 30th, the model starts to show abnormal current speed, negative temperature along the east boundary.

I don't expect to have any variation in temperature. I prefer it to be a constant.

So I have changed the LtracerSrc == F T .

Then the model blows up quickly and shows that I am getting negative temperature right at the river discharge point.

I will appreciate any ideas regarding this weird situation.

[kate]

This should be in the FAQ, it comes up so frequently. The river sources must have tracers set or else the model will go unstable. The only time you can turn off river tracers is for rivers leaving the domain.

So, yeah, we have to invent tracer values for some domains.

[txu68]

Hi Kate,

Thank you for your prompt response!

I was using LtracerSrc = T T, if this is what you mean by "must have tracers set".

But I got abnormal temperature, current speed along an open east boundary, which is nowhere near the river discharge location. And I started to get this abnormal thing after 30 days simulation, eventually this led to a blowup. Also, everything looks fine to me inside the river.

Do you have any suggestions about fixing the temperature along the open boundary?

Best,
Amy

[kate]

Code: Select all

LtracerSrc = T T

Yes, that's what I meant. So you had this before, then changed it and the behavior got worse. If it was blowing up at an open boundary before, what are your open boundary conditions?

[txu68]

These are my open boundary conditions.

Code: Select all

   LBC(isFsur) ==   Cha     Cha     Cha     Cha         ! free-surface
   LBC(isUbar) ==   Fla     Fla     Fla     Fla         ! 2D U-momentum
   LBC(isVbar) ==   Fla     Fla     Fla     Fla         ! 2D V-momentum
   LBC(isUvel) ==   Gra     Gra     Gra     Gra         ! 3D U-momentum
   LBC(isVvel) ==   Gra     Gra     Gra     Gra         ! 3D V-momentum
   LBC(isMtke) ==   Gra     Gra     Gra     Gra         ! mixing TKE

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

Abnormality starts to show up along the east open boundary, and it spreads into the interior domain. Eventually the temperature becomes negative and current speeds difference between the surface and the bottom is huge (~3m/s).

These are the cpp options.

Code: Select all

#define ROMS_MODEL
#define RAMP_TIDES
#define WET_DRY
#define UV_ADV
#define UV_COR
#undef  SALINITY
#define CURVGRID
#define SOLVE3D
#define MASKING
#undef   AVERAGES
#ifdef AVERAGES
 #define AVERAGES_AKV
 #define AVERAGES_AKT
 #define AVERAGES_AKS
#endif
#define STATIONS
#define DJ_GRADPS
#define SPLINES
#define TS_SVADVECTION
#define TS_U3HADVECTION
/*#define TS_MPDATA*/
#define ANA_PSOURCE
#define ANA_INITIAL
#define ANA_SMFLUX
#define ANA_STFLUX
#define ANA_BTFLUX
#define ANA_SSFLUX
#define ANA_BSFLUX
#define ANA_SPFLUX
#define ANA_BPFLUX
/*#define WESTERN_WALL
#undef EAST_M2REDUCED
#undef OBC_FS
#define EAST_FSCHAPMAN
#define EAST_M2FLATHER
#define EAST_M3GRADIENT
#define EAST_TGRADIENT
#undef  NORTH_FSGRADIENT
#undef  NORTH_M2GRADIENT
#define NORTH_FSCHAPMAN
#define NORTH_M2FLATHER
#define NORTH_M3GRADIENT
#define NORTH_TGRADIENT
#undef  SOUTH_FSGRADIENT
#undef  SOUTH_M2GRADIENT
#define SOUTH_FSCHAPMAN
#define SOUTH_M2FLATHER
#define SOUTH_M3GRADIENT
#define SOUTH_TGRADIENT */
#define SSH_TIDES
#define UV_TIDES
#define ANA_M2OBC
#define ANA_FSOBC
/*  #define ANA_TOBC*/
#define GLS_MIXING
#define KANTHA_CLAYSON
#define N2S2_HORAVG
#undef   SEDIMENT
#undef   SUSPLOAD
#undef   BEDLOAD
#undef   ANA_SEDIMENT
#undef   BED_CLASS
#undef  UV_LOGDRAG
#define UV_QDRAG
#undef UV_LDRAG
#define UV_DRAG_GRID
#undef SSW_BBL
#ifdef SSW_BBL
 #define SSW_CALC_ZNOT
 #define SSW_LOGINT
 #undef SSW_Z0RIP
#endif
#undef SG_BBL
#ifdef SG_BBL
 #define SG_CALC_ZNOT
 #define SG_LOGINT
#endif
#undef MB_BBL
#ifdef MB_BBL
 #define MB_CALC_ZNOT
 #undef  MB_Z0BIO
 #undef  MB_Z0BL
 #undef  MB_Z0RIP
#endif
#undef  ANA_WWAVE
#undef  SWAN

Best,
Amy

[kate]

#define EAST_FSCHAPMAN
#define EAST_M2FLATHER
#define EAST_M3GRADIENT
#define EAST_TGRADIENT

These will all be ignored by the modern model because it is reading that information from the LBC variables. Still, either way, you have a gradient condition. I don't know what sort of forcing you are giving it, but if you drive an inflow on any open boundary, the gradient condition will be just as badly behaved as your river source without any specified temperature, for exactly the same reason. If you want the temperature to be fixed, fix it on the boundary with your temperature boundary condition.

[txu68]

I am running a pure tidal simulation. So the only input forcing that I am giving is the tide.nc in different tidal components.

Everything else is using a default analytical solution from ROMS. In this case, where should I go to fix the temperature? Do you think it is the temperature that causes abnormal current that leads to the blow up?

Plus, I saw in ana_psource.h that we have Northeast_tile and Northwest_tile? What does those mean? Is it related to what I am facing here?

Best,
Amy

[kate]

Yes, it is probably the temperature that is causing trouble. Just use the clamped boundary conditions and set ana_tobc.h to what you want. The story with tracers is that on inflow, not clamping the values is going to give an unstable "downwind advection" scheme.