Difference between revisions of "Using the UMUI"

From UKCA
(create new page.)
 
Line 86: Line 86:
 
##*#* To allow the specification of these gases in this way to be uses by UKCA, you will also need to turn on this option in the UMUI: '''UMUI values for CFCs etc''' <br/> Model Selection -> Atmosphere -> Model Configuration -> UKCA Chemistry and Aerosols -> STRAT
 
##*#* To allow the specification of these gases in this way to be uses by UKCA, you will also need to turn on this option in the UMUI: '''UMUI values for CFCs etc''' <br/> Model Selection -> Atmosphere -> Model Configuration -> UKCA Chemistry and Aerosols -> STRAT
   
===UMUI UKCA Panel====
+
===UMUI UKCA Panel===
   
 
The UKCA configuration panel can be found at:
 
The UKCA configuration panel can be found at:

Revision as of 08:53, 7 September 2012

Required UMUI Changes

Before submitting a UKCA job, you must first alter several options in the UMUI:

  • User-id, Mail-id (email address), and Tic-Code:
    Model Selection -> User Information and Target Machine -> General Details
    The user-id is your username on the target machine (i.e. HECToR).
    The tic-code will be the HECToR charging code, e.g. n02-chem. This code can be found by logging into the HECToR SAFE pages, and viewing your account information.
  • Target machine root extract directory (UM_ROUTDIR):
    Model Selection -> FCM Configuration -> FCM Extract and Build directories and Output levels
    This needs to be the full path to one of your directories on the target machine (HECToR /home directory or MONSooN /projects directory).

Optional UMUI Changes

As well as the required changes above, you may also want to change the following, depending on your experiments:

  • Model start date and run length:
    Model Selection -> Input/Output Control and Resources -> Start Date and Run Length Options
    NOTE: For Stratospheric Chemistry, the CFC and other long-lived trace gas concentrations will be highly dependent on the date - you will need to think about your chemical tracer initial conditions if you change the start date of a Stratospheric Chemistry run.
  • Reference date for meaning:
    Model Selection -> Atmosphere -> Control -> Post processing, Dumping & Meaning -> Dumping and meaning
    If you have changed the start date (above), you may need to change the reference date. If you do not do this, climate meaning may not work. Climate means will only be produced for a date after the reference date.
  • Archiving:
    Model Selection -> Post Processing -> Main Switch + General Questions
    Currently archiving is switched ON for most jobs.
    On HECToR all model output will placed in a directory called archive which is a sub-directory of your run directory.
    ON MONSooN, all data is copied to /nerc/PROJECT/USERID/JOBID.
    Superseded model restart dumps will be deleted.
    You will also need to ensure that your Monsoon project group name is correct.
  • UKCA Surface Emissions (as user single-level ancillary file):
    Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies and potential climatologies -> User single-level ancillary files and fields
    The emissions data-sets are taken from those used for AR5 integrations, and represent the year 2000. You may wish to change these emissions - to do this we recommend the use of Xancil, provided by NCAS-CMS.
  • UKCA 3D Emissions (as user multi-level ancillary file):
    Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies and potential climatologies -> User multi-level ancillary files and fields
    The standard aircraft emissions data-set comes from the Aero2K project, and represents the year 2000. You may wish to change these emissions - to do this we recommend the use of Xancil, provided by NCAS-CMS.
  • Tracer Initial Conditions:
    Model Selection -> Atmosphere -> STASH -> Initialisation of User Prognostics
    The table in this UMUI panel gives a listing of all the initial conditions of the STASH items defined by the user. They are given a STASH code, which the UM uses to identify the field. The UKCA tracers have a 5 digit code beginning at 34001 and ending at 34150, although 34149 and 34150 need to be initialised to zero. Several (or all) tracers will be initialised to some initial conditions, contained in the file listed by each tracer.

Continuation Run

After the job has run the first step, you will probably wish to turn on a continuation run.

  • Turn off compilation:
    Model Selection -> Compilations and Modifications -> Compile options for the UM model
    Model Selection -> Compilations and Modifications -> Compile options for the UM reconfiguration
    Set to Run from existing executable, as named below in both panels
  • Set to CRUN:
    Model Selection -> Input/Output Control and Resources -> User hand edit files
    Find the hand-edit crun.ed and set this to Y

Running a Perpetual Year

For some experiments, you may wish to run a perpetual year (i.e. a simulation of multiple years in length, each one representing the same year). There are two ways to two this:

  1. Manually re-run the same year
    • In this case you would run up until you have the 1st January start dump, then manually restart the simulation from this dump, run for 1 year and then send a new job off, running from the new 1st January start dump. You would need to ensure that the date used for each of these steps is for the year that you are interested in. To do this you will need to change how reconfiguration is performed.
    • Reconfiguration:
      Model Selection -> Atmosphere -> Ancillary and input data files -> Start Dump
      You will need to alter the directory path and name of file to point to the start-dump you are interested in. You must then turn off the radio button Using the reconfiguation and make sure that the start-date (listed at the top of this panel) matches that in the dump file.
  2. Create a multi-year simulation where all the input files are for a single year
    • This method is particularly useful when you want to run a perpetual year for more than a few years, as might be the case for Stratospheric Chemistry when 15-20 year spin-ups are needed. There are a few changes that need to be made to several ancillary files and the UMUI:
    1. Change Ancillary files:
      • The following acillary files will need to be changed, since these are not climatologies by default, but time-series. You may find that the directory that these files are placed in is set by an environment variable (e.g. $ANCIL_ATMOS etc.). This environment variable points to a directory on the target machine (HECToR or MONSooN). To find out what these variables are set to, go to:
        Model Selection -> Input/Output Control and Resources -> Time Convention and SCRIPT Environment Variables.
        • The single-level and multi-level (see above) user ancillary files may need to be changed to the year you are interested in.
        • Sea-Surface Temperature:
          Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies & potential climatologies -> Sea surface temperatures
        • Sea-Ice:
          Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies & potential climatologies -> Sea ice fields
        • 2D Sulphur Emissions:
          Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies & potential climatologies -> Sulphur cycle emissions (2D)
          NOTE: you should not need to change the DMS emissions
        • Soot Emissions:
          Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies & potential climatologies -> Soot Emissions
        • Biomass Emissions:
          Model Selection -> Atmosphere -> Ancillary and input data files -> Climatologies & potential climatologies -> Biomass Emissions
      • To create perpetual-year ancillary files:
        1. For each of the the ancillary files specified in these UMUI panels above, open in Xconv and select only the year that you are interested in and export this as a NetCDF file.
        2. Use Xancil to create replacement ancillary files for each of the above files. You will need to make sure that these are made as periodic in time.
        3. In the UMUI, replace the time-series ancillary files with your climatologies.
    2. UKCA changes:
      These changes are describing code and UMUI changes specific to the branch fcm:um_br/dev/luke/VN7.3_UKCA_Chem.
      • The changes to UKCA are controlled by two hand-edits:
        Model Selection -> Input/Output Control and Resources -> User hand edit files
        1. CheM_aerosol_background_on.ed/CheM_aerosol_background_off.ed
          • This hand-edit either specifies using a background aerosol climatology for all years (_on), or using a time-series of atmospheric aerosols (_off). The default used is CheM_aerosol_background_off.ed.
        2. CheM_trgas_presentday_full.ed (a pre-industrial version is also supplied, CheM_trgas_preI_full.ed)
          • This hand-edit gives values of certain trace-gases to UKCA. Some are taken from the UMUI:
            Model Selection -> Atmosphere -> Scientific Parameters and Sections -> Spec of Trace Gases
            These are:
            • ()
            • ()
            • ()
            • ()
          • Some gases are used by UKCA to specify constant values for the chemical solver.
            These are:
            • (TropIsop chemistry only)
          • Some trace-gases are used in Stratospheric Chemistry, to specify the surface concentration of certain UKCA species (, , , , , and ). Some species are lumped together, so the surface concentrations of the following species can be set by the hand-edit:
            • (as MeBr)
            • (as MeCl)
            • (as MeCCl3)
          • A program to determine the mass mixing ratios of these species routine can be found at /home/ukca/bin/scenario on PUMA. It is run by typing

            scenario YYYY/MM/DD <which scenario>

            on the command line, where
            • <which scenario> can either be WMOA1, RCP2.6, RCP4.5, RCP6.0 or RCP8.5
            • The year (YYYY) is between 1950 and 2100 for the WMOA1 scenario and 1765-2500 for any of the RCP scenarios
            • The code uses a 360 day calendar (12 months, 30 days per month).
          • To allow the specification of these gases in this way to be uses by UKCA, you will also need to turn on this option in the UMUI: UMUI values for CFCs etc
            Model Selection -> Atmosphere -> Model Configuration -> UKCA Chemistry and Aerosols -> STRAT

UMUI UKCA Panel

The UKCA configuration panel can be found at:

Model Selection -> Atmosphere -> Model Configuration -> UKCA Chemistry and Aerosols

The settings in the UMUI UKCA panel have been chosen for that specific chemistry/aerosol scheme, and choosing other options may not work. If you require extra functionality, please contact Luke Abraham to check on the current code status.