Troubleshooting UV interactions w/ Nes boundaries

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
tviti3

Troubleshooting UV interactions w/ Nes boundaries

#1 Unread post by tviti3 »

Hi folks,

I have been playing around with synthetic models in ROMS, in order to try and build familiarity with the software and tooling, and recently encountered an issue when trying to add multiple refinement levels to one of my models. The problem is that the baroclinic velocities at the innermost grid's boundaries look like they are actually being influenced by said boundaries.

I've attached an animation of u_eastward at the surface, which illustrates this (the effect seems most pronounced on the N & S boundaries). Also attached is an animation of zeta, which at a glance doesn't seem to have this problem (although overall, there's not a whole lot of action in zeta).

u_eastward:
analysis--u-anim.gif
elevation:
analysis--elev-anim.gif
The model is a synthetic gaussian seamount, forced w/ prescribed tidal elevations. zeta at the east and west boundaries is 0.1m and 0.5m (resp.), oscillating at the M2 freq., while the north and south boundaries are just a linear interpolation between the east and west (all elevations are in-phase). To simplify things, temperature is the only tracer activated, with an arbitrary initial profile that has a marginal resemblance to that from a paper I am following (which the model geometry is also taken from).

The two "nested" grids are generated using the ROMS-matlab toolbox. Since I'm just using synthetic bathymetry, after each call to `coarse2fine`, I re-compute the bathymetry for the resultant refined grid, so that the refined grids are also using refined bathymetry. For the outer-most grid, dx = dy = 10km, which is refined around the seamount by 1:3, and then again by 1:3, so that on the innermost grid, dx = dy ~= 1.1km.

Some things I've tried playing with some options for the velocity and tracer mixing schemes (MIX_GEO_{UV,TS}, MIX_S_{UV,T}), and also tried using the coarse bathymetry (from the top-level grid) for all of the refined grids, neither of which seemed to eliminate this.

Thanks in advance

my_application.h:

Code: Select all

#define UV_ADV
#define UV_QDRAG
#define UV_VIS2
#define MIX_S_UV
#define DJ_GRADPS
#define SPLINES_VDIFF
#define SPLINES_VVISC
#define TS_DIF2
#define MIX_GEO_TS
#define SOLVE3D
#define FSOBC_REDUCED
#define RAMP_TIDES

#define NESTING

/* Analytical expressions */
#define ANA_INITIAL
#define ANA_SMFLUX
#define ANA_STFLUX
#define ANA_BTFLUX
#define ANA_FSOBC
#define ANA_M2OBC
Attachments
my_application.in
(142.33 KiB) Downloaded 169 times

Post Reply