An analysis is presented of tropical cyclone position and intensity forecasts issued operationally within the Australian region during the period 1973 to 1980. It is found that although the average errors were large the accuracy in movement forecasts steadily increased during the study period. Similarly, intensity forecasts were found to exhibit large errors. A diagnostic study of the ability of statistical objective schemes to predict tropical cyclone movement is also undertaken. Even when storm motion is difficult to forecast, past track data most often explain more variance than the synoptic data. However, the latter tend to become a more important predictive source for longer-term forecasts. A brief outline of research with important operational applications for the future is also given.
T. D. Keenan (Mon,) studied this question.