Senate debates
Wednesday, 10 October 2012
Questions on Notice
Bureau of Meteorology (Question No. 1849 amended)
Eric Abetz (Tasmania, Liberal Party, Shadow Minister for Employment and Workplace Relations) Share this | Link to this | Hansard source
asked the Minister representing the Minister for Sustainability, Environment, Water, Population and Communities, in writing, on 17 May 2012:
With reference to forecasts made by the Bureau of Meteorology (BoM):
(1) For the first three months of 2012, did BoM forecast:
(a) that most of eastern Australia, including all of Victoria and New South Wales, would exceed the median maximum temperature;
(b) that only a small part of south Western Australia bordering the Indian Ocean would have a significantly lower than average maximum temperature;
(c) that minimum temperatures across all of northern Australia and Western Australia would be higher than average;
(d) that all of South Australia, more than half of Queensland and Victoria, and approximately half of New South Wales would have below average rainfall; and
(e) higher than average rainfall in Western Australia.
(2) For each of the above paragraphs, from (a) to (e), what do the actual recordings now indicate.
Stephen Conroy (Victoria, Australian Labor Party, Deputy Leader of the Government in the Senate) Share this | Link to this | Hansard source
The Minister for Sustainability, Environment, Water, Population and Communities has provided the following answer to the honourable senator's question:
(1) The seasonal outlooks published by the Bureau of Meteorology provide what are known as a "probablilistic" forecasts. They do not forecast a definite outcome, but rather the chances or odds that a given area will be hotter or cooler, or wetter or drier, than the median value. They also do not predict how much above or below median.
Various measures of skill or accuracy can be used to assess a probabilistic forecast. In general, these measures compare the most likely (ie probabilistic) outcome to the subsequently observed outcome, and in turn how this would compare to a simple forecast scheme such as using climatology.
Seasonal forecast models are best evaluated over a large number of forecasts rather than using single events. They have proven to be valuable for decision-making when used with other relevant information as part of an assessment of risks.
Outlooks have demonstrated historical skill, with seasonal outlooks for temperature particularly accurate – for example in the last 12 months these show positive skill in all but 1 month across Australia.
The following responses are provided with reference to the seasonal outlooks issued in December 2011 for the first three months of 2012. The outlooks were for a:
(a) 50 to 80 per cent chance of exceeding the median maximum temperature over eastern Australia (or a 20 to 50 per cent chance of being below median);
(b) 50 to 75 per cent chance of being below the median maximum temperature for the western two-thirds of Western Australia;
(c) 60 to 85 per cent chance of above median minimum temperatures over northern Australia and Western Australia;
(d) 50 to 70 per cent chance of rainfall below the median across South Australia, the western half of New South Wales, and Queensland with the exception of the southeast quarter of the state; and
(e) 50 to 70 per cent chance of above median rainfall totals across Western Australia, southeast Queensland and the majority of eastern New South Wales.
(2) For each of the above paragraphs, from (a) to (e), the recorded seasonal conditions were:
(a) most of the country, including the east, recorded below-median maximum temperatures for this period;
(b) aside from the far west and a patch in the south, below median maximum temperatures were observed;
(c) seasonal minimum temperatures were mainly below the median except for patches of the north and the far west of Western Australia;
(d) above median rainfall was generally observed; and
(e) rainfall was above the median across most of Western Australia with below median observed in the far southwest.