Opened 4 months ago

Last modified 4 months ago

#2491 new help

North/South halos too small for advection

Reported by: charlie Owned by: um_support
Priority: normal Component: UM Model
Keywords: Cc:
Platform: NEXCS UM Version: 10.7

Description

Hi,

Following on from ticket number 2464, and the apparently well-known error "North/South? halos too small for advection" (see below for the whole error). I now know that it's not because of my Eocene-related changes (i.e. to the ancillary files) in u-aw739, because I have another suite (u-ay314) which is a modern run (i.e. no modifications at all to any of the ancillary files) which has also fallen over at exactly the same point and is giving exactly the same error:

???!!!???!!!???!!!???!!!???!!! ERROR ???!!!???!!!???!!!???!!!???!!!
? Error code: 15
? Error from routine: LOCATE_HDPS
? Error message: North/South? halos too small for advection.
? See the following URL for more information:
? https://code.metoffice.gov.uk/trac/um/wiki/KnownUMFailurePoints
? Error from processor: 666
? Error number: 19
????????????????????????????????????????????????????????????????????????????????

The only thing which links these 2 suites is the changeover of the aerosols, which I had to make mid-run from timeseries to 12 monthly climatologies (see comments 71 onwards in ticket 2464 for details). So do you think this is likely to be the cause?

Charlie

Change History (2)

comment:1 Changed 4 months ago by grenville

Charlie

How did u-aw739 behave - I'm guessing it doesn't suffer the same problem? Are you in contact with Will Roberts (owner of u-au884)?

Grenville

comment:2 Changed 4 months ago by charlie

Yes, it does suffer exactly the same error. Sorry if I wasn't clear. Basically I have 2 suites: u-aw739 (which has had all sorts of changes made to the ancillary files to make it Eocene, as well as a modified SST field and increased CO2) and u-ay314 (which is a straightforward modern version i.e. using all the original unmodified ancillary files and with everything else set to modern values). Both ran perfectly well for 20 years, until it came to the point where 3 of the aerosols ran out of time (because they were timeseries and only went up to 2010). So I changed these aerosols to the 12 monthly climatology versions, in both suites, and restarted. Both then failed one cycle further on, giving the same error about the halos. So surely it has to be because of this aerosol changeover? Luke warned that the changeover would create a spike (or dip, depending on the trend) in the relevant aerosols, but we thought that this impact would have worn off during my remaining 30 years of simulation. But perhaps it's this spike/dip that is causing the unrealistically high wind speeds, and therefore the error?

Charlie

PS. And yes, I am in regular contact with Will, but I don't think he has ever seen this error before.

Note: See TracTickets for help on using tickets.