GA 7.1 n216�Scott Wales & Holger Wolff
ARCCSS CMS
Aim:
Reproduce a current high-resolution climate run from the Met Office, as exactly as possible
Uses:
Daily mean temp + wind, 1950-01-05, data courtesy UKMO
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated
Source Job & Porting
Based on the GA7.1 HighResMIP run by the Hadley Centre, uses the faster EasyAerosol scheme
NCI site information is in the GA7.1 trunk, porting to Raijin involved minor changes and tuning the decomposition
Model runs at 28 hours per model year, costing 11 kSU per model year
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated
AMIP
One model year complete, scientific validation in progress by Peter Gibson (ARCCSS/UNSW)
Once match with UKMO results is good we will progress to JULES and CABLE bug fixes, running these for longer periods
The aim of these tests is to improve model results for both JULES and the UM, and to show the effectiveness of CABLE
Rose ID - u-an493
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated
Previous Test
How do we check two model runs with different compilers / libraries are giving the same results?
Run an ensemble with input fields perturbed by 1 ULP for each compiler, check if the results are sampled from the same distribution
Developed by João Teixeira at UKMO
UKMO
NCI
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated
Transpose AMIP
Initialise the model from ERA-Interim results at an arbitrary start date
Builds on work done at BoM
Allows for high-resolution modelling of the processes around extreme events
Rose ID - u-ap508
Precipitation and Surface Pressure, Brisbane floods Dec. 2010/2011
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated
More Info
ARCCSS CMS - climate_help@nci.org.au
Technical release only - not validated