FORECASTING ANNUAL DISCHARGE OF RIVER MURRAY, AUSTRALIA, FROM A GEOPHYSICAL MODEL OF ENSO

Annual discharge (Q) in the largest river system in Australia, the Riv er Murray (including the extensive tributary network of the Darling Ri ver), is often inversely related to sea surface temperature (SST) anom alies in the eastern equatorial Pacific Ocean. Conditional probability tables were constructed, with annual natural Q of the Murray for the period 1891-1985 divided into three amount categories; SST values were also divided into three groups. These tables permit probabilities of Q falling in each of three discharge categories to be estimated from e ither observed or forecast SST values. Using forecasts from a geophysi cal model, which indicated higher-than-average SST for most of calenda r year 1991, natural Q of the River Murray from June 1991 to May 1992 is forecast to be in the lower half of annual discharges since 1891 (6 4% probability). Using similar assumptions, the probability of annual natural Q for the year beginning June 1991 falling in the highest one- third discharge category is only 21%.