Gridded operational consensus forecasts

Abstract

The Gridded Operational Consensus Forecast (Gridded OCF) system is comprised of a "poor man's ensemble" of local and international numerical weather prediction (NWP) model forecasts of surface and upper air parameters such as temperature, dew point temperature, wind speed and wind direction, amongst others. These NWP forecasts are first bias corrected, and then composited by taking a skill-based weighted average of each bias corrected NWP forecast. Previous verification studies have shown that the skill of these consensus forecasts is better than the skill of any individual component NWP model. Gridded OCF has been used to provide guidance for the Bureau of Meteorology's official weather forecasts since 2008. Since then a number of important upgrades and improvements have been made to the operational system. The latest upgrade at the end of 2013 included increasing the spatial resolution from 0.5º to approximately 0.25º, producing forecasts on an irregular network of points rather than a regular grid, and the introduction of "intelligent grid point selection" for interpolating model data near coastlines. This poster compares verification results for the upgraded system with the previous operational system. It shows that the upgrades have resulted in significantly improved skill for near-surface temperature forecasts in mountainous areas and some coastal regions. Finally, suggestions for future improvements to the Gridded OCF system are presented, along with discussion on its utility for producing guidance for extreme weather events.