outflow at the open boundary condition

[winterbellgirl]

Hi all,

I'm running a model for the Gulf of Mexico. Now I'm trying to set up the open boundary conditions. The model has two open boundaries, the southern boundary opened at Yucatan and the eastern boundary opened at Florida strait. Since now I'm mainly focusing on the circulation of northwestern Gulf over the continental shelf, I do not expect the model to give a good simulation of the Loop Current, and the open boundary mainly serves at an outlet for water input by the Mississippi-Atchafalaya river discharge. My boundary configurations are as follows:

#define EAST_FSCHAPMAN
#define EAST_M2FLATHER
#define EAST_M3RADIATION
#define EAST_TRADIATION

#define SOUTH_FSCHAPMAN
#define SOUTH_M2FLATHER
#define SOUTH_M3RADIATION
#define SOUTH_TRADIATION


#define ANA_FSOBC
#define ANA_M2OBC

#define TCLM_NUDGING
#define TCLIMATOLOGY
#define M3CLM_NUDGING
#define M3CLIMATOLOGY

I modified the ana_nudgcoef so that nudging only applied to seven layers from the boundary to interior, both for T and M2. Temp and salt are nudged to climatogical values, while u and v are nudged to zero. Surface elevation and ubar/vbar at the open boundary use the default analytical values (zero). I ran the model for one year and the surface elevation in the interior increases with time. I guess this is because there is not enough outflow at the open boundary to output the water mass added by the river flows. Any solution for this? Thanks!

[simion1232006]

First thing is why you want to nudge your baroclinic velocity to zero. I think radiation for u , v should be enough.
Also, i had the similar problems when I ran a model with open boundary with no boundary value provided(you put it as zero).The flow can hardly cross the boundary when i use flather for M2 with zero barotropic velocity .
The flather boundary, I think, can preserve the volume if you do not have the the river discharge. There might be a imbalance of volume since you have a river discharge.So, maybe you can try radiation boundary for M2 to see whether it can dump all the inflow.

[rduran]

there is something called VOLCON (volume conservation) I believe there is info somewhere in the forum for VOLCON (and bry conditions associated with it) .. or was the info in the .in file? I'm not sure

[winterbellgirl]

Hi Simon,

Thank you for your suggestion for using the radiation boundary for M2. Actually when I increased the integration time, the surface elevation began to stabilize in the 2nd year, although the values are still high compared to observations (but the variability matches the obs well).So I guess I can use the current configuration. I used M3 nudging just to further reduce the instability at the boundary. Thank you.

[winterbellgirl]

Hi rduran,

I did see some discussions on VOLCON before. But i remember that somebody mentioned using VOLCON would increase the computational time by several times. This could be an alternative if there are no other solutions. Thank you for your hint!

[arango]

No, that is not true. The volume conservation is not expensive at all. I don't recall reading a message with such statement.

[winterbellgirl]

Thanks arango, and then I can try with the volcon.