Ocean and Marine Forecasting Group

Group Leader: N. Smith

Objectives: To improve understanding of the Australian marine environment, and to develop systems, including coupled ocean-atmosphere models, to predict ocean variations on time scales from days to decades.

Research activities

Climate research and prediction using coupled models

A new version of the coupled model (BCM4) is under development. Work for this model has focussed on the separate components. The base code for BAM4, the atmospheric component of BCM4, was released at the end of 2003. Basic testing of the climate configuration in AMIP style mode is being done in collaboration with other groups. The variability in the climate version of BCM4 is being analysed and a new version of the OASIS coupler is being implemented in BAM4. BMRC convened a workshop to discuss the development of a new Australian ocean model for climate research. The main outcome was an agreement for joint development of the Australian Climate Ocean Model (AusCOM). It will be the workhorse for the majority of climate prediction and predictability studies in the next 5-10 years.

The ocean and large scale atmospheric structure over the tropics and Australia during the 2002/3 and 1997/8 El Nino events have been compared using the NCEP re-analysis. There were large scale differences in the topical circulation which may be related to the location of the SST maxima in the tropical Pacific.

The investigation into the potential benefits of using Monte-Carlo forecasts to estimate error covariances for data assimilation was completed.

The POAMA-1 forecasting system continues to perform well. The National Climate Centre (NCC) assessment of coupled model forecasts shows that the POAMA performance is as good as, if not superior to, other models. The POAMA forecasts were the first to predict that the 2002/3 El Nino would decay rapidly at the beginning of 2003. The POAMA web site has been enhanced to include intra-seasonal forecast products. It is now widely used nationally and internationally for access to POAMA forecasts. Several papers have been prepared on the POAMA-1 system.

Ocean analysis and prediction

The BLUElink> Project, launched by the Parliamentary Secretary in October 2003 is a joint Bureau, CSIRO and Royal Australian Navy initiative based on the original Ocean Modelling and Analysis System (OMAS) project. The main Bureau of Meteorology commitments were: development of a data and product server; development of a high-resolution blended SST product and an enhanced surface wind analysis and forecast; development of enhanced ocean analysis and data assimilation systems; contributions to the development and testing of the ocean model, including improved ocean model physics; and ocean model forecasts, including intercomparisons with other products and studies of predictability.

Ocean data from a variety of sources is now being ingested in MARS. Procedures to store ocean thermal data have been developed and tested and various SST data products (non-Bureau sourced) are being archived. Existing systems will be changed to the MARS data source when MARS is upgraded. Work has commenced on the data and product viewer within AIFS. The Distributed Ocean Data System (DODS) server was commissioned in September 2003. Currently, POAMA results and some forcing fields are made available through the server.

A comprehensive review of available data and products and of possible techniques for blending SST data has been completed. Access to ENVISAT and AMSR-E SST data has been arranged (both to be stored in MARS) and negotiations are continuing for trial access to GOES 10 data. 9 km global AVHRR data will be used as the backbone for the global analysis. Local retrievals from 4 sites will be used as the basis for the regionally enhanced products. Research has begun on a possible technique for blending AVHRR, AATSR and microwave data.

Evaluation of the Bureau’s operational surface wind products has continued. Current work involves extending the previous verifications to more recent time periods. In addition, the feasibility of implementing an operational surface wind verification suite on a daily basis is being examined.

The MOM4 release of the ocean model has been installed at BMRC. Work has commenced on the configuration of this model. Work has also commenced on the distributed model intercomparison and evaluation framework, with up to four Centres likely to provide products in the Australian region.

A scientific planning meeting of the Antarctic Climate and Ecosystem (ACE) Cooperative Research Centre (CRC) was held in December 2003 to plan the contributions and strategy for each Program. Within Program 1, BMRC has developed its plans for a sea-ice analysis system and provided an outline of how it expected, with the Tasmanian Partnership for Advanced Computing (TPAC) to develop the sea-ice model assimilation system.

Ocean observations and data networks

There has been close collaboration with CSIRO Marine Research (CMR) on the deployment of Argo floats, including those floats purchased by the Bureau. The real-time handling of Australian float data (communicated through Argos) has been tested in parallel trials. Several weaknesses have been identified. Through the Joint Australian Facility for Ocean Observing Systems (JAFOOS), we have played an active role in coordinating Bureau of Meteorology and CSIRO ocean observation work programs.

Scientific quality control of delayed mode upper ocean thermal data from the Indian Ocean has continued. We continue to provide both quality assurance and scientific quality control of the Bureau's low density expendable bathythermograph (XBT) network dataset. Through the Navy project, a system for near real-time quality control is being trialled so that the ocean and climate prediction systems can have available good quality data from the region.

The end-to-end ocean data management project was completed in 2003 and produced a detailed analysis of the ocean and marine data systems within the Bureau and an upgrade of the SitesDB system. A draft Strategic Plan for Data Management was produced and a Workshop convened in November 2003 to discuss the Plan.

The Global Ocean Data Assimilation Experiment (GODAE) Project Office continues to function in support of various activities of GODAE. There were also many contributions to Bureau, national and international activities. These included participation in the Oceans Policy Science Advisory Group, responsibility as National Representative to the Intergovernmental Oceanographic Commission (IOC), and participation in development of the Bureau’s Ocean Services Program.

Marine forecasting

The storm surge model is ready for implementation and has been tested for some recent events (e.g. TC Monty in Western Australia in 2004) and made available to the Regional Offices.

Work on the development of an improved wave model background error correlation matrix has continued. A proposed new structure for the background errors that varies with latitude has been tested. Compared with the current operational system, it has been shown that the new structure can improve the skill of the wave model by approximately 10 per cent. Positive impacts were particularly strong for the central and north-east Pacific and for the Australian region.

The operational wave data assimilation routines were amended to address a limiting factor that prevented the wave spectrum from being adjusted in some situations in which the Significant Wave Height (SWH) was strongly overestimated by the wave model. This amendment resulted in a marginal improvement in the SWH verification statistics.

A review of potential parameterisations for the two-way coupling between ocean wave and atmospheric models has been undertaken. In collaboration with the Data Assimilation Group, the assessment of the accuracy of the operational surface wind fields has continued. Ongoing comparisons of LAPS and GASP 10m wind velocities with QuikScat scatterometer observations has shown that that over the last year, there has been a trend towards less negative bias in the modelled winds.