[MITgcm-support] Identifying location from processor & i,j

[Stanislav Martyanov]

Hello, Dafydd,

Some time ago I had a similar problem with the thickness of the upper
layer, though I do not use z* now. The cause of it was (in my opinion) the
seaice model. Nevertheless, in my run there was not multi-year accumulation
of sea ice, and also there were no thin bays in the domain (no 1-2 cell
thickness, I checked). Thanks to Martin's suggestion, the modification of
two parameters helped to solve the problem in data.seaice:

SEAICEnonLinIterMax = 10,
LSR_ERROR = 1.e-5

After this the problem vanished, though the changing of the ice loading can
also affect the problem. In my case, I changed the initial code in
seaice_growth and SEAICE.h, to work this HEFF rather than iceload. In fact,
it really affects only 4-5 cells in my domain, so it is not a problem.

Best wishes,
Stanislav Martyanov.


пн, 25 окт. 2021 г. в 22:53, Dafydd Stephenson <dafydd at ucar.edu>:

> Hi everyone,
>
> all of my runs seem to eventually (after a few model decades) exit with
> the error “ STOP in CALC_R_STAR : too SMALL rStarFac[C,W,S] ! ”. I am
> running the ECCOv4r4 configuration with CORE Normal Year Forcing.
>
>
>
> I have enabled monitoring at the timestep frequency, and, as the forcing
> repeats yearly, comparing the monitoring statistics with those of the
> previous year seems useful. Plotting them shows that they are comparable
> until the point of the error (at which point one of the line plots stops
> suddenly while the other continues), so I don’t believe there to be a CFL
> violation or any sort of runaway instability. Similarly, I have
> balanceEmPmR=.TRUE., so I don’t think that a cell is running dry.
>
>
>
> Nevertheless, the error seems to occur in the same location every time
> (processor 94 of 192 [or 47 of 96, depending on the config], at i,j=2,13),
> and I think it would be useful to know exactly where this is. Is there some
> way I could get from this CPU,j,i location to an index of the llc90 grid?
>
>
>
> Thanks in advance for any help,
>
> Dafydd
>
>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20211026/b36b1517/attachment-0001.html>