A comparison of five flood rain events over the New South Wales north coast and a case study

Abstract

One of the most significant forecasting problems facing meteorologists in the Sydney office of the Australian Bureau of Meteorology is the prediction of the heavy to extreme rainfall amounts that occur on the subtropical north coast of New South Wales (NSW). The significance of the problem lies in the great intensity of the rainfall, the difficulty in predicting the formation, location and duration of the rainfall, and the impacts of flooding on both cities and communities. The forecasting difficulties arise because of the small scales often involved, the range of physical mechanisms responsible and the challenge involved in forecasting events of such short duration. In the case of flash floods, the intense rainfall is typically less than 6 h in duration. Operational numerical weather prediction (NWP) has not shown much success, mainly because of the reasons mentioned above, especially given that computational limitations restrict its model resolutions. In this paper, the synoptic situation is presented for the most severe flash flood on record to affect Coffs Harbour, a city which is located on the NSW mid-north coast. The event, which occurred on the evening of 23 November 1996, is rated approximately as a 1 in a 100-year intensity/duration event. A 25-year climatology subset of flood-producing systems that have affected Coffs Harbour is then presented, and is correlated with the Southern Oscillation Index (SOI). Finally, details are given for a numerical modelling case study of the flash flood. The model results reveal that a very accurate prediction of the amount and the spatial distribution of rainfall amounts can be made for an event of this type. The rainfall totals of over 500 mm in 24 h were accurately forecast using coarse resolution archived operational initial analyses. The success of the predictions is most likely owing to the dominance of large-scale dynamical processes, which were further focused by the steep topography. Copyright © 2000 Royal Meteorological Society