The performance of objective tropical cyclone forecasting schemes in Australia is examined on limited samples of ‘best track’ and ‘operational track’ independent test data. The mean forecast errors indicate that persistence forecasts were generally more accurate although in some cases the differences were not statistically significant. Biases associated with each technique are documented although, with the exception of NHC-67, they are found to be a relatively minor error component. Simple combinations of the individual forecasts are shown to produce lower errors, although the consistent use of persistence within the techniques limits the overall gain of such approaches.
T. D. Keenan (Tue,) studied this question.