< back to Publications index

CSIRO Marine Laboratories Report Series

[back]

CSIRO Marine Laboratories Report 240

A New Version of the Australian Community Ocean Model for Seasonal Climate Prediction.

A Schiller*, J S Godfrey*, P C Mcintosh*, G Meyers*, N R Smith#, O Alves#, G Wang#, R Fiedler*

*CSIRO Marine Research
# Bureau of Meteorology Research Centre, Melbourne

Abstract: A model-to-data intercomparison is performed as part of the LWRRDC-funded project "Extended seasonal climate predications using a dynamical model". The Australian Community Ocean Model Version 2 (ACOM2) is the ocean component of a new seasonal prediction system for Australia, developed jointly between CSIRO Marine Research and the Bureau of Meteorology Research Centre. The main purpose of this report is to validate ACOM2 with observations. Furthermore, it is expected that the results presented in this report will serve as a reference for future work with this model. The report focuses on physical aspects of the model and addresses their potential impact on the forecast skill of the coupled ocean-atmosphere model.

A result of ACOM2 common to many other ocean general circulation models (OGCMs) is that simulated quantities representing the depth-integration circulation, such as depth-integrated steric height, sea-surface height or depth of 20 C isotherm are better simulated than surface quantities like e.g. SST. The near-surface circulation of the ocean is determined by model physics and exterior forcing (surface head and freshwater fluxed, wind stresses). Both of these components contain errors irrespective of the source of the surface forcing (from observations as used in this report or from an atmospheric circulatino model in a coupled mode), which hampers identification and rectification of model errors. Nevertheless, results presented here for oceanic surface circulation still reveal improvements over earlier models.

The main results of this report are summarised at the end. It is concluded that the winds used to drive the model seem to have serious deficiencies near 10 N in the Pacific. Most other differences between model and observation may be due either to model error or to flux problems. However, model error seesm a likely cause for the broad-scale overestimate of thermocline depth, and for misrepresentation of mean water temperature and salinity in the eastern equatorial Pacific.

Overall, the implemtation of an improved mixed-layer model, the parameterisation of tidal mixing in the Indonesian Archipelago, the enchanced merdional grid resolution, the better resolution of the Indonesion Archipelago and the reduced horizontal mixing all contribute to an improved ocean/atmosphere simulation in the coupled model runs, as was demonstrated in the comparison with the early version of the ocean model. None-the-less there is still space for significant improvements in the coupled ocean-atmosphere system, including data assimilation. This work will be a focus of the two modelling groups at BMRC and CMR over the next years.

Title Page (PDF 12kb)   Full Text (PDF 2.2Mb)

Updated: 12/11/08