Sudden temperature changes in the Sydney Basin: climatology and case studies during the Olympic months of September and October

Abstract

The accurate prediction of sudden large changes in the maximum temperature from one day to the next remains one of the major challenges for operational forecasters. It is probably the meteorological parameter most commonly verified and used as a measure of the skill of a meteorological service and one that is immediately evident to the general public. Marked temperature changes over a short period of time have widespread social, economic, health and safety effects on the community. The first part of this paper describes a 40-year climatology for Sydney, Australia, of sudden temperature rises and falls, defined as maximum temperature changes of 5°C or more from one day to the next, for the months of September and October. The nature of the forecasting challenge during the period of the Olympic and Paralympic Games to be held in Sydney in the year 2000 will be described as a special application. The international importance of the accurate prediction of all types of significant weather phenomena during this period has been recognized by the World Meteorological Organisation's Commission for Atmospheric Science. The first World Weather Research Program forecast demonstration project is to be established in the Sydney Office of the Bureau of Meteorology over this period in order to test the ability of existing systems to predict such phenomena. The second part of this study investigates two case studies from the Olympic months in which there were both abrupt temperature rises and falls over a 4-day interval. Currently available high resolution numerical weather prediction systems are found to have significant skill several days ahead in predicting a large amount of the detail of these events, provided they are run at an appropriate resolution. The limitations of these systems are also discussed, with areas requiring further development being identified if the desired levels of accuracy of predictions are to be reliably delivered. Differences between the predictability of sudden temperature rises and sudden temperature falls are also explored. Copyright © 2000 Royal Meteorological Society