[Mitgcm-support] Re: obcs in r1b1

[mitgcm-support at dev.mitgcm.org]

Hi,

the following lines cause the crash in orlanski_north.F (south, east and 
west similar):

          DO I=1-Olx,sNx+Olx
                ...
                OBNu(I,K,bi,bj)=uVel(I,J_obc,K,bi,bj)-
      &           CVEL_UN(I,K,bi,bj)*(deltaT/dyU(I,J_obc,bi,bj))*
      &           (ab1*(uVel(I,J_obc,K,bi,bj)-uVel(I,J_obc-1,K,bi,bj)) +
      &           ab2*(UN_STORE_4(I,K,bi,bj)-UN_STORE_1(I,K,bi,bj)))


where dyU(I,J_obc,bi,bj) is zero for i=-2 (Olx=3). To the best of my 
knowledge everthing should work as well limiting the loop to

          DO I=1-min(2,Olx),sNx+min(2,Olx)

In that case the radiation conditions work.

-Arne




Arne Biastoch wrote:
> 
> Hi,
> 
> try to use obcs in r1b1 and encountered a problem.
> When I run verification experiment 4 with the following data.obcs 
> everythings works fine:
> 
> # Open-boundaries
>  &OBCS_PARM01
>  OB_Jnorth=80*-1,
>  OB_Jsouth=80*1,
>  OB_Ieast=42*-1,
>  OB_Iwest=42*1,
> #useOrlanskiNorth=.TRUE.,
> #useOrlanskiSouth=.TRUE.,
> #useOrlanskiEast=.TRUE.,
> #useOrlanskiWest=.FALSE.,
>  &
> 
> But switching to pure Orlanski radiation conditions, i.e.:
> 
> # Open-boundaries
>  &OBCS_PARM01
>  OB_Jnorth=80*-1,
>  OB_Jsouth=80*1,
>  OB_Ieast=42*-1,
>  OB_Iwest=42*1,
>  useOrlanskiNorth=.TRUE.,
>  useOrlanskiSouth=.TRUE.,
>  useOrlanskiEast=.TRUE.,
>  useOrlanskiWest=.FALSE.,
> 
> 
> causes a core dump (floating divide by zero).
> Does that mean Orlanski radiation conditions are not working? Or am I 
> wrong?
> 
> -Arne
> 
> 
> 



-- 

Dr. Arne Biastoch

Institute for Marine Research         phone: ++49 431 600-4013
FB1 Ocean Circulation and Climate     fax  : ++49 431 600-1515
Duesternbrooker Weg 20                email: abiastoch at ifm.uni-kiel.de
24105 Kiel, Germany

http://www.ifm.uni-kiel.de/fb/fb1/tm/data/pers/abiastoch.html