Open boundary issues in NBQ simulations

Hello,

Here is a long-standing issue I have been experiencing with NBQ simulations. I am working with a domain that is the open on four sides. When the NBQ option is turned on, artificial surface waves tend to crop up and eventually lead to a ‘blow up’. I have attached two plots of the sea-surface elevation from a demo simulation in which the sea surface goes up and down like a seesaw to drive an oscillatory flow in the west-east direction. The water column is linearly stratified. As the simulation progresses, artificial surface waves appear (compare the 240 sec plot with the 960 sec plot), which tend to grow with time.

I suspect that this issue is due to the NBQ open boundary conditions. I find that the issue somehow worsens when I use OBC_NBQORLANSKI instead of OBC_NBQSPECIFIED.

My CPP options are as follows:
289 MPI
290 NBQ
291 OBC_EAST
292 OBC_WEST
293 OBC_NORTH
294 OBC_SOUTH
295 NEW_S_COORD
296 SOLVE3D
297 UV_ADV
298 SALINITY
299 NONLIN_EOS
300 UV_VIS2
301 UV_VIS_SMAGO
302 TS_HADV_WENO5
303 TS_DIF2
304 TS_VADV_WENO5
305 FRC_BRY
306 Z_FRC_BRY
307 M2_FRC_BRY
308 M3_FRC_BRY
309 T_FRC_BRY
310 ANA_BSFLUX
311 ANA_BTFLUX
312 OBC_M2CHARACT
313 OBC_M3ORLANSKI
314 OBC_TORLANSKI
315 BIO_HADV_WENO5
316 ANA_SMFLUX
317 ANA_STFLUX
318 NO_FRCFILE
319 DIAG_CFL
320 UV_HADV_C6
321 UV_VADV_C2
322 W_HADV_WENO5
323 W_VADV_WENO5
324 ANA_SRFLUX
325 ANA_SSFLUX
326 TS_MIX_GEO
327 UV_MIX_GEO
328 UV_VIS_SMAGO_3D
329 M2CHARACT
330 M3ORLANSKI
331 M2FILTER_NONE
332 NBQ_PERF
333 M3FAST
334 NBQ_IMP
335 NBQ_THETAIMP
336 HZR
337 NBQ_GRID_SLOW
338 NBQ_HZCORRECT
339 OBC_NBQ
340 OBC_NBQSPECIFIED
341 OBC_WORLANSKI
342 NBQ_NUDGING
343 NBQCLIMATOLOGY
344 NBQ_FRC_BRY
345 W_FRC_BRY
346 VAR_RHO_2D
347 TS_DIF_SMAGO
348 VIS_COEF_3D
349 DIF_COEF_3D
350 TEMPERATURE
351 NTRA_T3DMIX
352 NF_CLOBBER
353 TRACERS

Any suggestions on how to troubleshoot this issue will be much appreciated.

Best,
Guangyu


xupeng66
if you activate nudging it will improve your results? when there have Radiative conditions.
I don’t know that’s why i am asking.

Smaishal, thanks for your reply. Activating nudging does not lead to any improvement. Also, if I switch from specified to radiation boundary condition (e.g., OBC_NBQORLANSKI), the issue appears to worsen as the simulation blows up more quickly.