[MITgcm-support] OBCS for deep sewage water source

[Martin Losch]

Andrea,
sorry for the delay, I didn't get around to doing anything last week.  
And now I only have a short answer: I don't think that you can use OBCS  
in the way you want to. You have a source in the middle of the domain  
at the bottom, and I don't see why or how OBCS should work in this  
case. If I were you, I would prescribe a source at one grid point in  
ptracers_forcing.F (in tracer units/time), --- and if you want to  
simulate outflow through a sewer at the same point, some velocity  
source in external_forcing.F (S/R external_forcing_u), may some  
restoring--- and NOT use OBCS for the sewer problem. OBCS was not made  
for this type of situation, I guess. Any other opinions, anyone?

Martin

On Dec 13, 2004, at 4:14 PM, cimatori at libero.it wrote:

> Hi Martin, some time ago we wrote about this case: we are trying to  
> model the diffusion of a deep
> source of wastewater in a semi-enclosed basin (gulf of Trieste,  
> northern Adriatic) using obcs
> prescribe to attain the effect of lighter water, ptracers (for  
> modeling the polluttant) and exf. We
> use a western obcs at the open side of the gulf (we tried both  
> Orlanski and sponge (nudging) layer)
> and tried to use an eastern boundary for the sewage source. This  
> source is a diffuser offshore on
> the seabed, so we need to prescribe the boundary values only at a  
> particular level (15) leaving the
> rest of the column without any constraint.
> This works only if we use a closed domain (without any kind of western  
> open boundary): we see a
> plume of lighter water wich follows the wind-driven currents.
> When we open the western boundary we obtain irrealistic velocity  
> fields on the column above the
> diffuser.
> We corrected the routines (see attachments) making some improvements  
> but the main problem
> remains.
> The problem lies in the use of another open boundary together with the  
> one in the middle of the
> domain (we are going to try using northern or southern obcs instead of  
> eastern for the source:
> maybe it could be a C-grid problem?).
> We will check again the source code (expecially our modifications),  
> but, of course, any further
> suggestion is welcome!
> Thanks!
> Andrea
>
> PS: the passive tracer with obcs seems to give sensible results, at  
> least qualitatively, without
> using the deep source of light water (using only ptracer forcing). See  
> image attached.
>
>
>
> ____________________________________________________________
> Libero ADSL: 3 mesi gratis e navighi a 1.2 Mega, senza costi di  
> attivazione.
> Abbonati subito su http://www.libero.it
> <U_hor_sect_log(ptrac).png><obcs_apply_uv.F><obcs_apply_w.F>___________ 
> ____________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support