Opened 6 years ago

Closed 6 years ago

#1192 closed help (fixed)

HadGEM2ES - from couple run to amip run

Reported by: dh023729 Owned by: ros
Component: UM Model Keywords:
Cc: Platform: HECToR
UM Version: 6.6.3


Dear CMS,

After finishing a coupled version of HadGEM2-ES run, I begin to do a amip run. Switching off ocean coupling and replacing with sst and sea ice, I reconfigure the model and try to run it again.

This amip run (xhsfa) passes compilation but crashes in reconfiguration, with error:

 ERROR!!! in reconfiguration in routine Rcf_Calc_Len_Ancil
 Error Code:-  1
 Error Message:- StashCode is not a valid prognostic variable
 Error generated from processor  0

I've read a couple of previous tickets, and realise that this may be due to the mismatch of restartdump from coupled run and amip run, therefore, the restartdump from coupled run is replaced by a restartdump from amip run. However, the same error message is shown in the .leave file xhsfa000.xhsfa.d14009.t120226.leave.

Different from previous tickets, in this error message, the problematic STASH Code has not been given, no further information can be found in work directory neither. So, I am not clear which variable should be turn off from STASH.

I would be very grateful of any assistance in getting it work.


Change History (5)

comment:1 Changed 6 years ago by ros

  • Owner changed from um_support to ros
  • Status changed from new to accepted

Hi Liang,

Converting from a coupled job to AMIP is not trivial. Is there a reason why you can't start with the standard AMIP job (xgadd in the PUMA UMUI) and adjust it to your needs? Alternatively if a colleague has a more suitable AMIP job start could you start from that?


comment:2 Changed 6 years ago by dh023729

Dear Ros,

Thanks for the comment. I copied the AMIP job xgadd to my account (xhsfc), after changing userid, Tic code, Target machine root extract directory for FCM, I tried to do a test run without further modification.

However, the run failed at compilation (/home/n02/n02/dh023729/um/umui_out/xhsfc000.xhsfc.d14018.t153838.comp.leave). I scanned through this .leave file and could not find any error thought there are plenty of warnings. Would you please point it out for me what could be the reason for this failure?

Many thanks,

comment:3 Changed 6 years ago by ros

Hi Liang,

The compile was terminated because it ran out of time in the queue. It should have specified 1hour of time, but unfortunately due to a bug in a code fix that went in recently it didn't. I have now backed out the change so if you close and reopen your job in the UMUI and resubmit it, hopefully all will be well.


comment:4 Changed 6 years ago by dh023729

Dear Ros,

Thanks for the correction. And xhsfc is running.

Based on this benchmark run, further modifications are added, and of course, more errors appear.

Modifications I had put in are:

  1. Change Start Date from 1978/01/01 to 1870/01/01.
  2. Consequently, relevant ancillary files need to be replaced to contain data dated back to 1870/01/01. These ancillary files are included:
    sulphur (2D) emissions, soot emissions, biomass emissions and fossil-fuel organic carbon emissions.

However, error occurs as:

 UM ERROR (Model aborting) :
 Routine generating error: UP_ANCIL
 Error code:  427
 Error message:
REPLANCA: Current time precedes start time of data

I read through the subroutine REPLANCA, and realised that unless I modify this subroutine and get more information printed out, I can not figure out which ancillary field is field 27.

Any comment and assistence will be appreciated.


P.S.: .leave file is /home/n02/n02/dh023729/um/umui_out/xhsfc000.xhsfc.d14021.t123356.leave

comment:5 Changed 6 years ago by grenville

  • Resolution set to fixed
  • Status changed from accepted to closed


From the leave file:

20 Atmos & Slab Ancillaries to be updated.
Anc Ref No 7 Stash code 60 Interval 30 Days
Anc Ref No 27 Stash code 31 Interval 1 Days
Anc Ref No 28 Stash code 24 Interval 1 Days
Anc Ref No 29 Stash code 32 Interval 1 Days

Field 27 is stash 31, which is Fraction of sea ice - the ancillary data starts 1870/10/16 - so your model is trying to start too early.


Note: See TracTickets for help on using tickets.