I’m trying to adjust a suite (u-dd849) which previously ran with pre 2015 conditions to now continue running from 2015. I have changed the ancils and other necessary parts of the rose-app.conf but the suite keeps failing with the below message.
? Error code: 65536
? Error from routine: EM_GET_TIME_INFO
? Error message: NetCDF calendar 360_day does not match model calendar, and attribute calendar_flexible /= 1 or not set
? Error from processor: 0
? Error number: 69
As far as I can tell, all the emission files are either gregorian or are 360 day with calendar_flexible=1. I used the same set of emissions in an older gregorian calendar run I did under post 2105 conditions so I’m at a loss.
If a run is started at 1st Jan 2015, does anything else need to be done to allow for the transition?
This is an annoying message, as you have discovered, as it doesn’t give the file name! If you look in the pe0 file in the pe_output directory, does it give any more information? The usual suspects are the CH4 emissions for this error.
Another possibility is the offline oxidants or the easy aerosol files if these are used in the run. These are also read by the same routines.
Sorry, another issue has arisen in a related suite, u-dd879, where I am also trying to run post 2015 (the above calendar issue was present in this suite too but has been fixed)
I switched the ancils to post 2015 files but the model is failing on the first resubmission (April 2015) with
? Error code: 500
? Error from routine: UP_ANCIL
? Error message: REPLANCA: error finding next lookup entry for field: 4 stashcode 216
? Error from processor: 0
? Error number: 63
I have had this before and it is caused by ancils with the wrong dates (i.e pre 2015) and in the atmos_main/pe_output, the ancil file being read is indeed a pre-2015 file even those though the suite specifies a post-2015 file.
FILE_MANAGER: Assigned : /projects/ukesm/jwalton/ancil/u-by791/u-by791_m01s00i216_1979-2014_annual_timeseries_land_cover_frac.anc
FILE_MANAGER: : Unit : 37 (portio)
Ancil file num : 3
Ancil file unit num: 37
Ancil filename : /projects/ukesm/jwalton/ancil/u-by791/u-by791_m01s00i216_1979-2014_annual_timeseries_land_cover_frac.anc
Stash req = 216
The correct post-2015 emissions have been read in along with the SST and Sea Ice ancils but the other land ancils and ocean biogeochemistry ancils are also for pre-2015.
So it appears the suite is selecting some ancil files from rose-suite.conf and some from another source.
For some UKESM suites they were configured (very badly, in my opinion) to use an optional config file to define some file paths. This is particularly terrible for users, as it overwrites settings in the Rose panel with a file that is not necessarily easy to find. For things that don’t change, like compiler settings etc. this makes sense, but in my opinion, this should never be done for forcing files, as you might end up completely screwing-up someones experiment.
Please take a look and see if there exists a app/um/opt/rose-app-archer2.conf file. If so, this will likely be overwriting your settings. I would suggest removing all forcing files from it (update the settings in the main app/um/rose-app.conf file if needs be). Once that is done you will then be able to set files correctly in the main suite.
I’ve found a rose-app-monsoon.conf file in the opt subdirectory which was defining the ancils which were being overwritten in my suite. I have removed the files from this opt file and will run again.
Great! Sorry, I should have realised this was a Monsoon2 suite. The same has happened on ARCHER2.
If you haven’t changed these files in the main Rose suite you’ll need to have copied them across to the main file, otherwise the suite might crash as the original files might not exist (hence them being overwritten in the optional config file).
Sorry, a further confusing error has appeared at the start of the 2nd resubmission (July 2015).
targ_time,time1,time2 322500.00000 322512.00000 323244.00000
hours,int,period 298320 1 -1
Information used in checking ancillary data set: position of lookup table in dataset:21
Position of first lookup table referring to data type 1
Interval between lookup tables referring to data type 1 Number of steps20
STASH code in dataset 96 STASH code requested 96
‘start’ position of lookup tables for dataset 6in overall lookup array 84153
Information used in interpolation/replacement:
Time of first data=322512.00000
Validity Time for update=322500.00000
Time of second data=323244.00000
UP_ANCIL : Error in REPLANCA.
???
???!!!???!!!???!!!???!!!???!!! ERROR ???!!!???!!!???!!!???!!!???!!!
? Error code: 507
? Error from routine: UP_ANCIL
? Error message: REPLANCA: TIME INTERPOLATION ERROR
? Error from processor: 0
? Error number: 393
???
Stash code 96 is Chlorophyll and the ancil for that is a standard post-2015 file
Just an update on this. I tried running u-dd849 (a parallel suite to u-dd879) with the same ancils from 2015 but with 1 month resubmission (rather than 3 months u-dd879) and it failed midway through September with the same error, again linked to ancil 96.
I’m afraid that I haven’t seen this error before. I’m hoping someone else in CMS has and may be able to help. Otherwise, perhaps contact Jane at the Met Office and see if anyone there knows anything?