6A.1
Australian approaches to probabilistic precipitation forecasting

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 25 January 2011: 1:30 PM
Australian approaches to probabilistic precipitation forecasting
613/614 (Washington State Convention Center)
Michael Foley, Bureau of Meteorology, Darwin, Australia; and S. Cooper, P. Riley, E. Morgan, and J. Bally
Manuscript (589.7 kB)

Forecasting of rainfall probabilities and amounts has until recently received little emphasis in services provided by the Australian Bureau of Meteorology, despite the potential benefits of such information to users. The main product which has been made available for users since 2006 has been a fully-automated rainfall forecast using a “poor-man's ensemble” of NWP models to provide daily probabilities of exceeding certain rainfall thresholds, as well as a daily rainfall amount.

With the introduction of the US-developed Graphical Forecast Editor (GFE) into Australian forecast offices, commencing in 2008, forecasters have become directly involved in forecasting rainfall probability and amount. We will describe tools which have recently been introduced into operations in the Australian GFE to enable the meteorologist to produce daily forecasts of precipitation probability and amount in an efficient manner, making use of an assumed form for the cumulative rainfall probability distribution function. We will also describe more sophisticated approaches which are under development.

The “poor-man's ensemble” rainfall forecasts provide the main input for this forecast process. This guidance currently suffers from lack of calibration. We will present recent improvements which involve calibration using historical data.

The benefits of improved forecasts of rainfall probabilities and amounts will only be felt in the community if the new outputs are effectively incorporated into public forecast products. Plans for communicating information on rainfall probability and amount to the Australian public will be described.