Seasonal forecasts are central to good risk management in agriculture and other weather-sensitive industries as they enable informed planning and decision making well in advance of undertaking key activities. Skillful seasonal forecasts can be used to maximise benefit in good years as well as avoid losses in bad years. A basic question in forecasting is whether one prediction system is more skillful than another.

The seasonal forecast information is not complete if we have no idea of the forecast skill. Computing and providing skill measurements is a crucial step for the use and further development of seasonal forecast. Particularly when multiple seasonal forecasts are now freely available, and they may disagree.

Rainfall climatology during summer months (December-February) for observations (AWAP) and climate model (with 1-month lead time)

This study assesses the ability of 6 global dynamical climate models to forecast monthly rainfall in Australia:

• ECMWF (SEAS5)
• Météo-France (System 5)
• UK Met Office (GloSea5-GC2)
• DWD (GCFS2.0)
• NCEP (CFSv2)
• BOM (ACCESS-S1)

About the presentation:

This analysis was presented at the Australian Meteorological and Oceanographic Society Annual Meeting and the International Conference on Tropical Meteorology and Oceanography (AMOS-ICTMO 2019), held in Darwin from 11-14 June 2019. ACCESS-S1 was excluded from the slides because this model has been calibrated unlike the rest of the climate models.

Please, contact us if you want to know more results or would like to provide us any feedback.