Notes of the code management group meeting 21/3/2013

From UKCA
Jump to: navigation, search

Notes from code managment meeting held 21/3/2013 in Cambridge

This meeting covered two main issues: the release of a shared job at 8.4 and plans for lodging code in the next 12 months.

Getting to a shared release

  1. The prefered job will be a N96L85 job based on Global Atmosphere 4 (paper submitted to GMDD).
  2. It will use the CHEST chemistry scheme with the ASCHEM extention if possible and the stratospheric version of GLOMAP-Mode.
  3. The fallback if this does not work is the ACHEM chemistry and GLOMAP-mode without the aerosol extensions
  4. The CHET chemistry version can still be used for development but is not to be the released job
  5. The planned timeline is:
    1. End of May - configuration ready to be run for 1 year to generate input to evaluate suite.
    2. Mid June. Luke to do actual model run and run the evaluation suite.
    3. End of June - Alex to look at chemistry results from evaluation suite and Graham to look at aerosol results and use a Red-Amber-Green system to rate them. Red means it is not suitable for doing any science with at the present time. Amber means there are still some issues but if these are documented, they should not prevent its use for peer reiviewed publications. Green would be the best results we have so far seen from UKCA.

Code to lodge in next 12 months

To start the process of NCAS partners lodging code directly to the MetUM trunk, in the next year we will try and lodge at least one branch from Cambridge and one from Leeds. The following are candidates for lodging this year by NCAS.

  1. RCP scenario code. Luke Abraham to lodge.
  2. CHES+ code. Luke Abraham to lodge. Lower priority than RCP code
  3. Flightrack code. Nick to discuss with Paul Telford.
  4. Metzger scheme for organic mediated BL nucleation. Graham to try and lodge this.

Support to be provided by Met Office for this process will include:

  1. Nick to provide advice and training on setting up tickets and testing
  2. Met Office to set up UTF and UMDP3 compliance checker on postproc
  3. Nick to look into making Cambridge laptop a pool laptop
  4. If code is developed by someone who does not have a Met Office account, Met Office to find someone to do the final steps to lodge the code (which will only include final UTF testing and possibly upgrading 1 version, if code fails review it will go back to the original author)

In addition, code from Unviersity partners which will be lodged by Met Office staff include:

  1. Model dust
  2. Nitrate aerosol
  3. Convective plume scavenging

Leeds and the Met Office will work closely on the dust and nitrate schemes. For a more complete list see the UKCA trac pages.