Confusion about abnormal flow and great temperature gradient

[mengqingjun]

Hi everyboday, I am running a real application without tide forcing on East China Sea. The model resultion is 1/12 degree. I run my model for 10 years in two cases:
1.The model is forced only by the boundary file *_bry.nc and it is interpolated from the HYCOM 1/12 degree results. The variables in the boundary file only have 12 time records for 15,45,75,105,135,165,195,225,255,285,315,345. I also make the initial file interpolated from the HYCOM results. The LBC conditions I used are as follows:

Code: Select all

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

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

and the Nudging time scales I used are as folloows:

Code: Select all

       TNUDG == 2.0d0 2.0d0              ! days
       ZNUDG == 0.0d0                      ! days
      M2NUDG == 2.0d0                      ! days
      M3NUDG == 2.0d0                      ! days
OBCFAC == 60.0d0                      ! nondimensional

I specify my spatially varying nudging coefficients as follows:

Code: Select all

      cff1=1.0_r8/(30_r8*86400.0_r8) ! 30-day outer limit
      cff2=0.0_r8 ! Inf days (no nudge) inner limit
      cff3=20.0_r8 ! width of layer in grid points

! cff3-point wide linearly tapered nudging zone
      DO j=JstrR,MIN(INT(cff3),JendR) ! SOUTH boundary
        DO i=IstrR,IendR
          wrk(i,j)=cff2+(cff3-REAL(j,r8))*(cff1-cff2)/cff3
        END DO
      END DO
! cff3-point wide linearly tapered nudging zone
      DO i=IstrR,MIN(INT(cff3),IendR) ! WEST boundary
        DO j=JstrR,JendR
          wrk(i,j)=MAX(wrk(i,j), &
     & cff2+(cff3-REAL(i,r8))*(cff1-cff2)/cff3)
        END DO
      END DO

! cff3-point wide linearly tapered nudging zone
      DO i=MAX(IstrR,Lm(ng)+1-INT(cff3)),IendR ! EAST boundary
        DO j=JstrR,JendR
          wrk(i,j)=MAX(wrk(i,j), &
     & cff1+REAL(Lm(ng)+1-i,r8)*(cff2-cff1)/cff3)
        END DO
      END DO
!
! cff3-point wide linearly tapered nudging zone
      DO j=MAX(JstrR,Mm(ng)+1-INT(cff3)),JendR ! NORTH boundary
        DO i=IstrR,IendR
          wrk(i,j)=MAX(wrk(i,j), &
     & cff1+REAL(Mm(ng)+1-j,r8)*(cff2-cff1)/cff3)
        END DO
      END DO
!

my cpp definations are as follows:

Code: Select all

#define UV_ADV             
#define UV_COR             
#define UV_QDRAG           
#define UV_VIS2            
#define DJ_GRADPS           
#define MIX_S_UV
#define MIX_S_TS 
#define TS_DIF2           
#define TS_U3HADVECTION     
#define TS_C4VADVECTION 
#define SOLVE3D           
#define SALINITY          
#define NONLIN_EOS          
#define CURVGRID
#define PROFILE   
#define MASKING
#define AVERAGES
#define MY25_MIXING
#if defined MY25_MIXING
# define KANTHA_CLAYSON           
# define K_C2ADVECTION      
# define K_C4ADVECTION      
# define N2S2_HORAVG        
#endif
#define ANA_BTFLUX          
#define ANA_BSFLUX          
#define ANA_SMFLUX
#define ANA_STFLUX
#define ANA_SSFLUX

I make some month averaged plots using the 10th model year,but I find abnomal flow and great temperature gradient. I wander if there is something wrong with my boundary file or something wrong with my ROMS parameters in ocean.in file or something wrong with my CPP definations in oecan.h file ?

2.Base on case1, I only add the atmospheric forcing. I use

Code: Select all

#define QCORRECTION
#define SRELAXATION

But the abnormal flow and great temperature gradient are still exist. Why ? Anyone could kindly help me ? Thank you !

[rduran]

A monthly BC is not a good idea, imagine following scenario to illustrate: a relatively strong flow coming in, you have that more or less fixed for a period of time because of the low frequency BC, but your model has higher frequency stuff going on the interior and it wants flow to go out.

Strong flow in + strong flow out = strong and unreal vertical velocities through conservation of volume.

You may be seeing artificial and very strong upwelling due to something like this.

That would be my guess, but how come you have temperatures as low as 0.4? Isn't that's weird? I don't know the regional oceanography much to be able to know bottom waters ...

[kate]

Also, your nudging timescales are really very short. I usually set TNUDG and friends to 360, with OBCFAC of 120, for an inflow timescale of 3 days. Yours is quite a bit shorter than that, in fact your nudging on outflow is stronger than my nudging on inflow, which could be trouble.

As for the spatially varying nudging timescale code you show, I'm not sure it's being used for anything since you don't have TCLM_NUDGING or similar turned on. However, if you were to turn it on, the nudging timescales set here would trump the boundary ones from TNUDG in the default code.

[mengqingjun]

Thank you rduran for reply!

A monthly BC is not a good idea

,do you mean that a daily BC will do better for a ten year model ? Mybe I will try a daily BC...

how come you have temperatures as low as 0.4

I think the temperature could reach as low as 0.4, may even be negtive one...I think it is reasonable, is't it?
By the way, I think the temperature ROMS used is potential temperature, not the in-situ one, am I right ? Thank you!

[mengqingjun]

Thank you Kate for reply !

in fact your nudging on outflow is stronger than my nudging on inflow, which could be trouble

I see, I will try your's set.

I'm not sure it's being used for anything since you don't have TCLM_NUDGING or similar turned on

If I turn on TCLM_NUDGING or similar, ROMS will use the spatially varying nudging timescale code I set in ana_nudgcoef.h instead of TNUDG and friends I set in the ocean.in file, am I right? If I don't turn on TCLM_nudging or similar, ROMS will use the ones in the ocean.in file and the nudging will only have on the boundary, am I right ?
ps. If I try a daily BC, will TNUDG and friends to 360, with OBCFAC of 120, for an inflow timescale of 3 days be also OK ?

[kate]

Yes, that's correct. Subject to change without notice, of course. Yes, you can have a nudging timescale of 3 days to a field that changes daily, no problem.

[rduran]

do you mean that a daily BC will do better for a ten year model ? Mybe I will try a daily BC...

in my small amount of experience daily boundary conditions work very well. If you can get data at the same times when ROMS asks for a BC you might not even need nudging at the boundaries at all... that is how my current model is running. Now 10 years * 365 days maybe a lot of data ... I dont know ... you'll have to check if you can get that to work. Worst case scenario, note you can have several BC files ... so it sounds like you should be fine.

I think the temperature could reach as low as 0.4, may even be negative one...I think it is reasonable, is't it?

It could be, again it depends on your regional oceanography. I don't expect minus one in my region ... but not all regions are the same. ROMS does use potential temperature and in situ density. it is only potential temperature and salinity that are treated as conserved quantities (except when there are sources or sinks that is).