Skip to content

Commit

Permalink
Documentatino clean-ups. COSIMA/access-om2#109
Browse files Browse the repository at this point in the history
  • Loading branch information
Nic Hannah committed Sep 6, 2018
1 parent e8ad372 commit fad5ffb
Showing 1 changed file with 7 additions and 16 deletions.
23 changes: 7 additions & 16 deletions README.md
Expand Up @@ -26,22 +26,13 @@ The model tracks two datetime variables throughout a run:
Typically these two datetimes will have the same month, day, hour, minute, and second but may differ in the year. For example a Repeat Year Forcing experiment may start with experiment and forcing datetime of 1990-01-01T00:00:00 but as the run progresses to the second year the experiment datetime will be 1991 but the forcing datetime will go back to 1990.

Here is a description of all the namelist variables used to control datetime variables:
* forcing\_start\_datetime, forcing\_end\_datetime: the start and end of the forcing period. The model will be forced by a continuous repetition of the forcing period.
* restart\_period: interval of time between successive model restarts. This is provided as a tuple: years, months, seconds. This breaks the entire experiment into a collection of segments.

Other key variables in datetime management are:
* exp\_start\_datetime: the experiment start datetime. this is defined to be the same as the forcing\_period\_start\_datetime and cannot be set by the user.
* exp\_cur\_datetime, forcing\_cur\_datetime: the current experiment and forcing datetime. These are maintained by the model including across restarts and cannot be set by the user.
* dt: the timestep, this is not set by the user but is read from the forcing files.
* calendar: 'noleap' or 'gregorian' calendar, once again not set by the user but read from forcing files.

# What am I doing now
* forcing\_start\_datetime, forcing\_end\_datetime: the start and end of the forcing period. The model will be forced by a continuous repetition of the forcing period.
* restart\_period: interval of time between successive model restarts. This is provided as a tuple: years, months, seconds. This breaks the entire experiment into a collection of segments.

Running ACCESS-OM2 tests. Update payu for yatm. Add model level module, accessom2_mod. Set path to namcouple. implement accessom2%sync_config()
Other key variables in datetime management are:

CI tests (in priority):
- Reproducing previous results (e.g. compare against current ACCESS-OM2 output).
- Run with default settings
- RYF and IAF dates
- RYF and IAF restarts
- minimal fields - this looks tricky because the dest is masked.
* exp\_start\_datetime: the experiment start datetime. this is defined to be the same as the forcing\_period\_start\_datetime and cannot be set by the user.
* exp\_cur\_datetime, forcing\_cur\_datetime: the current experiment and forcing datetime. These are maintained by the model including across restarts and cannot be set by the user.
* dt: the timestep, this is not set by the user but is read from the forcing files.
* calendar: 'noleap' or 'gregorian' calendar, once again not set by the user but read from forcing files.

0 comments on commit fad5ffb

Please sign in to comment.