22 March 2013: Forecast from POAMA are now produced twice per week starting on Thursdays and Mondays. We are working on the web-site to take advantage of these more regular forecasts, e.g. Using a larger ensemble by combining previous forecasts to make a larger ensemble (e.g. For seasonal we go back two weeks to make 5x33=165 member ensemble). This leads to forecasts that are more reliable. The plots are gradually being updated and upgraded to take advantage of this lagged ensemble approach. Other enhancements will also be implemented in the next few weeks.
14 February 2013: Monthly and Seasonal forecasts using multiple forecast runs have been created. Changes to the format and composition of these forecasts will be ongoing.
28 January 2013: Started running a forecast on Mondays as well as Thursdays.
30 November 2012: POAMA to Applications Workshop 29th-30th November 2012 completed.
26 November 2012: New gridded format for the Rainfall and Temperature anomaly forecasts.
26 November 2012: Error corrected in rainfall and temperature anomaly forecasts. Forecasts had a 50% scale error, which was more evident for the Temperature forecasts than the Rainfall forecasts as the logarithmic scale minimises the effect of this error. This error was associated with the multi-week system (M24) products only.
01 October 2012: POAMA to Applications Workshop dates changed to 29th-30th November 2012.
24 September 2012: POAMA to Applications Workshop 1st-2nd November 2012.
24 September 2012: Introduced POAMA-2 seamless products span time scales from weeks to seasons (all forecasts are now based on the M24 system).
03 May 2012: Introduced 9 month lead times for Multiweek forecasts.
14 December 2011: Added Hindcast Plots to Multiweek MJO POAMA forecasts.
25 November 2011: Added pages for MJO using the Multiweek POAMA forecasts.
25 November 2011: Added pages to show the Multiweek Skill.
25 August 2011: POAMA realtime forecast are now password protected. POAMA hind-casts and forecasts are freely available in delayed mode (3 months behind real-time for forecasts from the seasonal and 6 weeks behind real-time for forecasts from the multiweek systems).
22 June 2011: Plan for Multiweek experimental products.
A monthly/multi-week version of POAMA-2 has been transferred to Bureau operations. Experimental/prototype products will become available during July and August. This version will produce a 30 member ensemble every Thursday and is more tailored towards the shorter multi-week timescales.
20 June 2011: Preliminary POAMA-2 Website goes live.
Preliminary version of experimental products web-site from POAMA-2 goes live. POAMA-2 is currently being run by Bureau operations in trial mode and is expected to officially replace POAMA-1.5 in late 2011.
20 June 2011: POAMA-1.5 Experimental Products Web-site Decommissioned.
POAMA-1.5 Experimental products web-site was decommissioned on 20 June, due to the server being closed down at the Bureau. Note POAMA-1.5 is still the operational model and operational products are still available here (http://poama.bom.gov.au/operational_products.shtml).
Questions about using the web-site
- Why is the latest forecast not available?
POAMA-2 is currently running in operational trial mode in parallel to POAMA-1.5. It does not receive 24hr support until it replaces POAMA-1.5 as the official system and therefore the running of the model is sometimes delayed with products only available after the web-site has been updated. If the latest forecast does not automatically load, use the navigation buttons to go back to the previous forecast.
- When will multi-week products be available from POAMA-2?
Experimental products from the multi-week system are now available.
Questions about POAMA
- Why is POAMA-2 showing a different forecast to POAMA-1.5, for example, the SST indices
relating to El Niño-Southern Oscillation (ENSO)?
POAMA-2 uses a new process for integrating ocean and atmospheric observations, which is a significant improvement over the system used in POAMA-1.5, particularly in the representation of ocean salinity. Testing of the model on past ENSO patterns has revealed that POAMA-2 is better at predicting ENSO events (ie. La Niña or El Niño) than POAMA-1.5. We therefore believe that POAMA-2 forecasts are more accurate than those from POAMA-1.5.
- Where can I find more information about POAMA-2?
More information regarding POAMA-2 can be found heres (http://poama.bom.gov.au/about_poama2.shtml). Additionally, scientific papers describing the new systems incorporated into POAMA-2 are listed on our Publications page (http://poama.bom.gov.au/publications.shtml). Further papers describing the system and its skill are being prepared and will be made available on our publications page in the near future.
- How skilful is POAMA-2?
Papers documenting the skill of POAMA-2 are being prepared and will be made available when ready. In general POAMA-2 shows increased skill for ENSO and local Australian quantities.
- Are there Skill plots for POAMA-2?
These are available on the web-site
- What is the difference between POAMA and ACCESS?
POAMA is seasonal prediction system, one component of which is a coupled ocean-atmosphere general circulation model (GCM), ACCESS is a national project that is developing a common modeling framework for a full spectrum of applications from weather forecasting to climate prediction and change, which includes the development of a new coupled GCM. POAMA-3 will utilise the ACCESS coupled model as its core modeling component.
- When will POAMA-2 become operational?
POAMA-2 will replace POAMA-1.5 (the current operational model) late in 2011. Official Bureau products currently using output from POAMA-1.5 (see here http://poama.bom.gov.au/operational_products.shtml) will continue to be available once the changeover occurs, but will instead use POAMA-2 output. This web-site (http://poama.bom.gov.au/poama2.shtml) introduces new experimental products from the POAMA-2 system. Subject to sufficient demand and resources, some of the POAMA-2 experimental products will become official Bureau products over the next 1-2 years.
- What is the difference between an official Bureau product and an experimental product?
Official or operational products are provided by the Bureau’s operational divisions and are fully supported by the Bureau and technically maintained 24 hours a day.
Experimental products are provided by the Centre for Weather and Climate Research (CAWCR) as research quality products. They do not currently form part of the Bureau's standard services in any way. Access to the products is made available for research purposes only and on the basis that users are fully aware that these products are still being tested. While these products are available in real-time this cannot be guaranteed as 24-hour operational support is not available.
- What is the difference between the multi-week and seasonal products?
The multi-week and seasonal products are generated from two different systems. The seasonal system is run on the 1st and 15th of each month and produces forecasts out to 9 months. The multi-week system is run every Thursday and produces forecasts out to 9 months. This system also incorporates a more sophisticated ensemble generation system that generates more reliable forecasts at shorter leads. Because of the overlap in lead times some products will be available from both systems.
- What is the difference between the forecast and the hindcast?
The forecast is based on the observations available at the time. After all data and observations have arrived, a clearer picture of the weather conditions can be derived and it is called a re-analysis. The hindcasts are based on the re-analysis files and these provide slightly different initial conditions to those that are used to create a forecast.
- Why are experimental products not of consistent presentation quality?
Experimental products are produced by different researchers from within CAWCR hence showing some stylistic variations. Being still in the research phase, there has not yet been any efforts to ensure continuity in their presentation. Additionally these products are generally targeted at researchers rather than end-users, and as such some products may show quantities that may not be meaningful to end-users, but help scientists to better understand the POAMA forecasts.
- How can I provide feedback on the POAMA-2 experimental web-site?
Feedback is always welcomed and can be provided via email addressed to: email@example.com
- I find the products useful and I’d like to start using them. When are they likely to
become official Bureau products?
There are plans to make some experimental products official Bureau products over the next two years. If there are particular products you’d like to see become operational you can communicate your interests via the Bureau’s feedback page http://www.bom.gov.au/other/feedback/.