spatial discontinuity problem

[ywang152]

Dear all,

I am working on a case study of simulating the water current of estuary without wind forcing. The boundary condition and initial condition (including temperature, salinity, ubar, vbar, uvel, vvel, zeta) were interpolated from another hydrodynamic model simulation result. I have complete a 72-hr simulation without crash. However, the problem is that the result shows spatial discontinuity as the attached gif file shown.

animation of temperature simulation reuslt

The boundary type I used is:
free surface: Chapman
2d momentum: Flather
3d momentum: Clamped
Tvar: Clamped
** I don't have MTKE on boundary.
I also test other kinds of boundary type combination, for example
1)free surface-Chapman, 2d momentum-Flather, 3d momentum-Radiation, Tvar-Radiation
2)free surface-Radiation, 2d momentum and 3d momentum-Clamped, Tvar-Radiation
3)all are Clamped
4)all are Radiation
They didn't help with this problem...

I suspect the boundary condition may have some problem, so I tried to do some idealized run: only open one boundary (north boundary) and only apply free surface as boundary condition(Clamped), and set the initial condition as 0. However, the simulated zeta of the cell inside of the domain has high frequency oscillation, which is pretty different with the water level on boundary. I also attached a figure to show the zeta time series of cell on the north boundary (blue line) and the cell inside of the domain (orange line). I guess the reason of the oscillation is I closed all the other boundaries, but I am not sure....

zeta time series of two cell, blue is for cell on north boundary, orange is for cell inside of the domain

I also attached the *.in file(ocean_estuarynw_p_mix.in) and standard output file (run_3.log).
If you have had similar problem or have ideas about my spatial discontinuity problem, thanks for your advice!

Best regards
Yifan