GENIE:GENIE Versions

=GENIE Versions (internal page)= This page records the details and meanings of tags in the GENIE code repository. Please also see the externally available page for a summary.

The GENIE code base is tagged at various junctures with the following naming convention:

rel---

HEAD
Note that you can think of the repository's HEAD version as a kind of model beta version. Any code commited to the repository should have been tested against the suite of short tests. These tests are designed to run easily and quickly--to encourage people to use them. They also aim to exercise as much of the model code in the short time that they take to run. However, you will appreciate that there is an inherent tradeof. To supplement the short tests, a longer suite of tests is run as part of the nightly build.

Released versions are scrutinised a good deal more than the automated tests can do. For example, the model is may have been subjected to a number of tuning runs, exercising the model across large portions of the space of tunable parameters. In addition to the elevated quality control, may also have a set of optimised parameters to use in conjunction with a stable version of the model. (There is no guarantee that such a parameter set will be optimal for another version of the model's source code.)

So, in summary, it is far better to use a released version of the model than the HEAD.

===Features which are currently planned to be added, in preparation or evaluation, which may already be available as an option in the trunk but are still to be evaluated, or which may be available in a branch and are planned to be merged into the trunk===
 * Various tuned configurations of,  , and
 * Configurations for different meshes: '3636s08l', '3636s16l', '7272s16l' (topography of '3636s16l'), '7272s32l' (topography of '3636s16l'), '3660l08l', '3660l16l', '6432i08l', '6432i16l' ('iijjGkkl' - longitudinal resolution of ii points, latitudinal resolution of jj points on grid type G (s: equal-area grid, l: equidistant spacing in longitude and latitude, i: igcm-matching grid), kk vertical layers in the ocean)
 * BIOGEM: Equilibrium scavenging of 231Pa, 230Th
 * BIOGEM: Optional separated biological production of siliceous and non-siliceous phytoplankton
 * GOLDSTEIN/EMBM: Seasonal wind forcing

rel-2-7-6

 * Release of version 2.7.6 (SVN revision 6270):
 * : run-time interpolation of data-based fields of PO4 and alkalinity used for RMS model-data difference computation
 * Extended centralised input and output functionality
 * Corrections and adjustments

rel-2-7-5

 * Release of version 2.7.5 (SVN revision 6067):
 * Corrections
 * Additional topography-related input files

rel-2-7-4

 * Release of version 2.7.4 (SVN revision 5902):
 * : run-time interpolation of data-based fields used for RMS model-data difference computation
 * genie-wind: optional wind speed and wind stress forcing fields based on run-time interpolation of a reanalysis dataset

release-2-7-3

 * Release of version 2.7.3 (SVN revision 5798):

rel-2-7-2

 * Release of version 2.7.1 (SVN revision 5767):
 * GOLDSTEINSEAICE: Explicit or implicit transport scheme can be selected via namelisted parameter

rel-2-7-1

 * Release of version 2.7.1 (SVN revision 5558): update to 2nd release within the scope of the GENIE-LAMP project
 * Corrected restart mechanism in ENTS module (SVN revision 5541)
 * Other updates
 * Base tag for tuning exercise Tunings of the versions 2.7.n

rel-2-7-0

 * Release of version 2.7.0 (SVN revision 5515): 2nd release within the scope of the GENIE-LAMP project
 * BIOGEM: fractionation of 30Si
 * Other updates
 * Base tag for tuning exercise Tunings of the versions 2.7.n

rel-2-6-0

 * Release of version 2.6.0 (SVN revision 5428): 1st release within the scope of the GENIE-LAMP project
 * Merge of code from  into   removes redundant code
 * Base tag for tuning exercise Tunings of the versions 2.6.n

rel-2-5-1

 * user.mak environment variables configured for iwan.ggy.bris.ac.uk environment. For UoB GENIE workshop.

rel-2-5-0

 * Tag immediately after merging of Martin and Greg branches.

rel-2-4-3

 * Stable ... I think! release for Darwin Center / Utretch GENIE modelling workshop.

rel-2-4-2

 * Stable release for UoB MSc course.

rel-2-4-1

 * Bug fix after development on trunk broke model.

rel-2-4-0

 * ENTS has been geniefied!

cup-1-0-0

 * Version 1.0 of cupcake. Merges to/from trunk to follow.

rel-2-3-0

 * Tag immediatly after switching to XML.

rel-2-2-1

 * Last tag before switching to XML format config files.

rel-2-2-0

 * Tag created for UEA workshop.

rel-2-1-0

 * GOLDSTEIN equation of state corrected
 * Corrections to embm, goldstein and seaice modules to align eb_go_gs with genie-cgoldstein

rel-2-0-0

 * Formerly tagged XmasGC
 * Version used for the "Grand Challenge" experiments conducted between Dec. 2005 - Mar. 2006

rel-2-7-5-tuning

 * Used for phase 2 of the tunings of the versions 2.7.n

rel-2-7-4-tuning

 * Used for phase 1 of the tunings of the versions 2.7.n

rel-2-7-1-tuning

 * Tunings of the versions 2.7.n
 * Changes made on branch merged into trunk version
 * Deleted, superseded by branch

rel-2-6-tuning

 * Tunings of the versions 2.6.n
 * Superseded by tuning exercise for versions 2.7.n

Marsh_et_al-2009-GMDD

 * Model version used for the revision of the manuscript by Marsh et al., Geosci. Model Dev. Discuss., 2, 1-79, 2009
 * Has been synchronised with trunk version at revision 5902 (rel-2-7-4) plus specific subsequent revisions made on the trunk and local modifications

Other Branches
Un-merged branches in the CVS repository for specialised study

ipcc-ar4-2005-rel-1-2

 * Corrected equation of state
 * Version used for final submission (Oct. 2005)

ipcc-ar4-2005-rel-1-1

 * Minor correction to inm.f
 * Version used for initial submission (Mar. 2005)

ipcc-ar4-2005-rel-1-0

 * Initial tag intended for production runs in the ipcc-ar4-2005 branch

Model Comparisons
On the 14/12/06 Gethin had the following final values for average SAT (reported to STDOUT):


 * {| border="1" cellpadding="3"

!Model !command !value
 * cgoldstein
 * "goldstein < goin.test")
 * 10.5088016763012
 * genie_eb_go_gs
 * "genie_example.job -f configs/genie_eb_go_gs.config"
 * 10.5088016337715
 * }
 * 10.5088016337715
 * }
 * }

On the 12/02/07 Gethin had the following final values for average SAT, after doubling the run length:


 * {| border="1" cellpadding="3"

!Model !command !value
 * cgoldstein
 * "goldstein < goin.test")
 * 11.3269478462944
 * genie_eb_go_gs
 * "genie_example.job -f configs/genie_eb_go_gs.config"
 * 11.3269577445993
 * }
 * 11.3269577445993
 * }
 * }