The large-scale dust event that was observed in much of central and eastern Australia over the period 22-24 October 2002 was one of the largest experienced in Australia in recent decades. Its large scale, and the high quality observations of the storm from a variety of observing platforms, provided the opportunity to verify forecasts of the dust transport during this event from two particulate transport models.

Analyses from the Bureau of Meteorology’s Limited Area Prediction Scheme were used as a diagnostic tool to describe the synoptic dynamics that allowed strong winds from aloft to reach the surface, where gusts > 40 kn persisting through the nocturnal hours were observed in post-frontal winds. Comparisons of output from both transport models with observations from satellite and ground-based platforms are presented at four times over a 36-hour period during the duststorms’ transit of mainland Australia. It is demonstrated that the performance of forecasts from both transport models in predicting the position of the dust front was quite good. However, it is shown that the forecasts are very sensitive to the land-surface specification that delineates potential source regions for the dust, and these data require further development before operational
dust forecasting using these models could be relied upon.