[MITgcm-support] Re: pickup problem

[Alistair Adcroft]

Tom,

Leith is state dependent and has no "save" information of it's own. If 
the pickup is failing it is because it would fail anyway (without 
Leith). I checked the data file and there is nothing specific to pickup 
version; the non-linear free surface needs special treatment but it 
should be doing that by default (Jean-Michel?).

Another possibility is some inconsistency with the forcing/exf stuff; 
I'm not sure how it works across pickup boundaries so need to defer to 
Patrick or Ed?

A.

Tom Haine wrote:
> Dear Alistair, 
> 
> I'm having trouble picking up my new spill jet calculation. The pickup
> causes a numerical instability within a few steps once the flow is more
> or less equilibrated. Specifically:
> 
> i) I stop at step 4320 and write a pickup dump. I restart at step 4320
> and get: 
> 
> WARNING: r*FacC < hFacInf at: i,j,bi,bj,Thid,Iter=  71  21   1   1   1
> 4323
> rStarFac,H,eta = -0.527735  5.000000E+00 -7.638676E+00
> STOP in CALC_R_STAR : too SMALL rStarFacC !
> 
> in STDERR.0032. I.e., the code dies 3 steps after pickup. This error
> message seems to be associated with divergence of the Leith scheme.
> 
> but:
> 
> ii) I can run from step 0 past step 4320 with no problems if I don't
> stop and pickup (I reach step 4608 and all is well before running out of
> time on the lsc).
> 
> This seems to point directly at a Leith pickup problem. What do you
> think?
> 
> Cheers, Tom.

-- 
Dr Alistair Adcroft              email: Alistair.Adcroft at noaa.gov
NOAA/GFDL, Princeton University               Tel: (609) 987-5073
Forrestal Campus, U.S. Route 1, P.O. Box 308, Princeton, NJ 08542