[MITgcm-support] [EXTERNAL] Re: rolling pickups: nchecklev value on restarting model

[Menemenlis, Dimitris (US 329B)]

Hi Dan, I am not sure if relevant to your question, but see:
http://wwwcvs.mitgcm.org/viewvc/MITgcm/MITgcm_contrib/high_res_cube/input/modpickup
https://github.com/MITgcm-contrib/ecco_darwin/blob/master/v03/cs510_latest/input/modpickup
for example scripts that automate the choice of latest available complete pickup files and continuation of simulation.

D.
________________________________________
From: MITgcm-support <mitgcm-support-bounces at mitgcm.org> on behalf of Jean-Michel Campin <jmc at mit.edu>
Sent: Friday, August 19, 2022 11:13:03 AM
To: mitgcm-support at mitgcm.org
Subject: [EXTERNAL] Re: [MITgcm-support] rolling pickups: nchecklev value on restarting model

Hi Dan,

I think that what you describe is accurate: these ckptA/ckptB restart files are "rolling"
pickup, so they are going to be over-written as the run progress. And the first rolling pickup
written (after starting the simulation at iter=0 or after any restart) is "ckptA", the second
is "ckptB", and then repeated.

One thing that has been added (a long time ago) is a short print statement to STDOUT
each time a pickup is written, that is easy to grep for:
> grep '%CHECKPOINT ' STDOUT.0000
and to get the last one:
> grep '%CHECKPOINT ' STDOUT.0000 | tail -n 1

Cheers,
Jean-Michel

On Thu, Aug 18, 2022 at 10:57:58AM +0100, Daniel Goldberg wrote:
> Hello All
>
> In the setup Im working with, it matters to me whether at a given time, ckptA or ckptB is the suffix used for a pickup write! There is a variable nCheckLev in a common block in RESTART.h which seems to determine this behaviour.
>
> >From grepping for ths variable, it seems to me that when we start from a pickup, this variable is reset.
>
> Thus, if ckptFreq is set to e.g. 10000., and i do a run that is 10,000 seconds long, then pickup.ckptA.* will be written at the end of the run, assuming deltaT divides into 10,000. If the run were 20,000 seconds or more, then pickup.ckptB.* would be written at 20,000 seconds -- here im just describing rolling pickups.
>
> BUT, assuming nCheckLev is reset each time the model is (re)started, then if I were to then restart the model from the pickup.ckptA.* written at 10,000 seconds, then pickup.ckptA.* would be OVERWRITTEN at 20,000 seconds.
>
> I wonder if someone could comment on whether this is correct?
>
> Many thanks
> Dan
>
>
>
> --
> Daniel Goldberg, PhD (he/him)
> Reader in Glaciology
> School of Geosciences, University of Edinburgh
> Geography Building, Drummond Street, Edinburgh EH8 9XP
>
>
> em: dan.goldberg at ed.ac.uk<mailto:dan.goldberg at ed.ac.uk>
> web: https://urldefense.us/v3/__https://www.geos.ed.ac.uk/homes/dgoldber__;!!PvBDto6Hs4WbVuu7!Z2nNI9cm1AhzNfWJOwa4uSOCJvqDQkoOa66sZGyS6DVmUPZ_ZNqMJoCTOZYtywBMb45YJ9Nq30g$
> The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. Is e buidheann carthannais a th??? ann an Oilthigh Dhùn Èideann, clàraichte an Alba, àireamh clàraidh SC005336.

> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> https://urldefense.us/v3/__http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support__;!!PvBDto6Hs4WbVuu7!Z2nNI9cm1AhzNfWJOwa4uSOCJvqDQkoOa66sZGyS6DVmUPZ_ZNqMJoCTOZYtywBMb45YwCiNOOA$

_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
https://urldefense.us/v3/__http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support__;!!PvBDto6Hs4WbVuu7!Z2nNI9cm1AhzNfWJOwa4uSOCJvqDQkoOa66sZGyS6DVmUPZ_ZNqMJoCTOZYtywBMb45YwCiNOOA$