Lengthwise instabilities in velocity leading to blow-up

[Landonky]

I am currently experimenting with ROMS on rivers, this means I'm likely running ROMS with on smaller time and length scales than are typically used.
This river has a mean depth around 5m, and a damn regulated flow around 200 m³/s.
I define v_south/north by dividing Q by the cross-sectional area. I set the outflow (NORTH) zeta to zero and calculate zeta_south by calculating a slope:

dz/dx = ( drag * u² ) / ( Hyd.Rad * g )

my time stepping is dt=0.05s (half of CFL), DTF=20, DT=1

I currently think my problem is in my number of vertical layers, the most I can get to run is 5.

The model will run fine up to 1.5hrs past the one hour ramp (see t=150min), and then these lengthwise velocity perturbations grow from the edges and propagate down the river. The models typically run the full 4hrs without blow up, but the results clearly are not right.

Is this a problem with too many layers? Too small an hmin? bad options? I pretty new at this so any insight or recommended articles are welcome!

I've included my model options and small snap shots of the inflow region.

-Kyle

Code: Select all

 NoCPP               KO-3D-07E
 ANA_BTFLUX          Analytical kinematic bottom temperature flux.
 ANA_INITIAL         Analytical initial conditions.
 ANA_SMFLUX          Analytical kinematic surface momentum flux.
 ANA_STFLUX          Analytical kinematic surface temperature flux.
 ASSUMED_SHAPE       Using assumed-shape arrays.
 COSINE2             Cosine-squared shape time-averaging barotropic filter.
 CURVGRID            Orthogonal curvilinear grid.
 DJ_GRADPS           Parabolic Splines density Jacobian (Shchepetkin, 2002).
 DOUBLE_PRECISION    Double precision arithmetic.
 EASTERN_WALL        Wall boundary at Eastern edge.
 GLS_MIXING          Generic Length-Scale turbulence closure.
 KANTHA_CLAYSON      Kantha and Clayson stability function formulation.
 MASKING             Land/Sea masking.
 MIX_S_UV            Mixing of momentum along constant S-surfaces.
 MPI                 MPI distributed-memory configuration.
 NONLINEAR           Nonlinear Model.
 !NONLIN_EOS         Linear Equation of State for seawater.
 NORTH_FSCHAPMAN     Northern edge, free-surface, Chapman condition.
 NORTH_M2FLATHER     Northern edge, 2D momentum, Flather condition.
 N2S2_HORAVG         Horizontal smoothing of buoyancy and shear.
 PROFILE             Time profiling activated .
 K_GSCHEME           Third-order upstream advection of TKE fields.
 !RST_SINGLE         Double precision fields in restart NetCDF file.
 SOLVE3D             Solving 3D Primitive Equations.
 SOUTH_FSCHAPMAN     Southern edge, free-surface, Chapman condition.
 SOUTH_M2FLATHER     Southern edge, 2D momentum, Flather condition.
 SPLINES             Conservative parabolic spline reconstruction.
 TS_C4HADVECTION     Fourth-order centered horizontal advection of tracers.
 TS_C4VADVECTION     Fourth-order centered vertical advection of tracers.
 UV_ADV              Advection of momentum.
 UV_U3HADVECTION     Third-order upstream horizontal advection of 3D momentum.
 UV_SADVECTION       Parabolic splines vertical advection of momentum.
 UV_QDRAG            Quadratic bottom stress.
 UV_VIS2             Harmonic mixing of momentum.
 VAR_RHO_2D          Variable density barotropic mode.
 WESTERN_WALL        Wall boundary at Western edge.
 WET_DRY             Wetting and drying activated.

[Landonky]

The problem was that I was forcing the inflow with a depth uniform velocity U=( Q(t) / A ). This was causing problems at the bottom and edges of my channel. To get around this I created a deep entrance to allow for slower velocities and blended it into the actual bathymetry.