Some aspects of the verification of weather forecasts for Melbourne, Australia

The authors have recently completed a piece of work exploring trends in the skill of weather prediction at lead times of 1 to 14 days for Melbourne, Australia. The system that was used to establish these trends (at longer lead times) was, in part, based upon an algorithm that statistically interpreted the GFS NWP model output to generate local weather forecasts. It was considered that it would be interesting to assess what might be achieved using the output of other global NWP models. Preliminary results are presented about what has been achieved using the ECMWF monthly control NWP model. The general public's first impression of the forecast weather is provided by the official précis of that forecast. Using an algorithm that interprets the words contained in the précis in terms of precipitation provides guidance on how successfully the précis provides useful information about precipitation amount and probability. Capability at predicting unusual weather events is of importance. Some statistics are presented on how effectively one can now predict weather that is somewhat different from climatological norms. Finally, the skill of the aforementioned system at predicting differences in temperatures recorded at the current Melbourne Central Business District (CBD) site, and those at a new observation site just outside the CBD, is discussed.