Location of output data /transferring to JASMIN


Just starting running UM – could anyone assist on the questions below?

  1. I have just run test job of 2 months HadGEM3 3.1 (u-cq957)
    Where should I look for output data files?

  2. Also I would like to transferring output to Jasmin:
    I have specified the jasmin directory on
    Post Processing -common settings
    ARCHER2-JASMIN Archiving
    Data Transfer
    Transfer_dir ‘/gws/pw/j05/arctic_connect/jgrist002/HG3_initial’

Are there any other things to change,
Many thanks,

Hi Jeremy,

I’m aware that you are looking for a CANARI-like suite, but just to say that this suite is very much still in active development and has quite a few extra bells & whistles specific to the project.

In answer to your questions:

  1. The suite outputs both Atmos & Ocean data in netcdf format using XIOS. The output ends up in the cylc-run/u-cq957/share/cycle/<cycle-name> directory.

  2. In order to use pptransfer to transfer data to JASMIN you will need to setup a Gridftp certificate as detailed here: Configuring PPTransfer


Hi Ros,

Many thanks for these answers,


Hi Ros,

Thanks for your help on those queries. We’d thought we’d explain a bit more about our goals.

Our objective is to run a series of 6 months runs starting from end of a spun up control run.

The scientific goal is to see how the atmosphere responds to changes in the ocean mixed layer.

There would be 2 experiments.

In the first we’d use the control run ocean/ice restart but different atmosphere restarts.

The second would be like the first but with the ocean restart perturbed by us offline.

We know how to change the ocean restart, but we don’t know where the atmosphere restarts can be specified.

(Could you point us to this).

We realise the CANARI suite maybe too experimental for us, could your recommend another a suite which already has a long control run from which we could start.

For example a CMIP6 pre-industrial control (or similar with the same N216 ORCA025)?

With many thanks for your help with this,


Hi Jeremy

CANARI is probably suitable - it’s still under development but is close to being production ready. We shall continue to work on the configuration. You might take a copy of the u-cn134 trunk. u-cm829 might be alternative but it is a UM 10.7 suite that will not be developed.

Typically, the atmosphere start file is specified in astart. If reconfiguration is required, then set ainitial.

Please consider attending the UM training coming up (Introduction to Unified Model - NCAS)


Hi Grenville,

Many thanks


This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.