My regional OM3 config crashes with the following error:
WARNING from PE 607: Extreme surface sfc_state detected: i= 211 j=1028 lon= 137.796 lat= -32.646 x= 137.796 y= -32.646 D= 4.1000E+00 SSH= 9.9130E+00 SST= 2.4501E+01 SSS= 4.5001E+01 U-=-2.6554E-02 U+= 0.0000E+00 V-= 6.8799E-02 V+= 0.0000E+00
FATAL from PE 30: There were a total of 1 locations detected with extreme surface values!
Salinity > 45 psu is the issue here (images show the bathy and sea surface salinity with the offending cell shown with an x). If unclear from the bad zoom in, this spencer gulf near Adelaide.
The 1 grid cell-wide section of this embayment seems sub-optimal, and I plan to edit it out.
But I also learned that Salinity ~ 45 psu is not unheard of here, and you can see Salinity approaching 45 psu seaward of the constriction. So, would this be a case where I would prescribe BAD_VAL_SSS_MAX = 50.0 or something? Are there any hidden dangers?
Yeah I reckon bumping up the max salinity would be best! I don’t expect anything too crazy to happen when bumping it up by just a bit. Maybe it was set with coarse open ocean conditions in mind, and a super salty surface right on the coast wasn’t considered
Hi @mmr0 - I have had similar issues around Spencer Gulf and bathymetry modifications were needed for me. If the salinity fix doesn’t do the job and you need to go down the bathymetry modification route then you may also need to regenerate one of the mesh files access-rom3-ESMFmesh.nc
Setting a minimum depth of 10 – 15 meters when generating the domain in the regional mom6 toolbox might also fix the issue.
I always ended up just filling in those kinds of embayments where there was a 1-2 gridcell wide channel, reasoning that the model isn’t going to be entirely useful in those little pockets anyway. Filling in is made very easy with those gridtools (thanks to @aekiss@angus-g and others who created them!)
