The Seasonal Climate Outlook for Pacific Island Countries (SCOPIC) is a statistical climate forecasting tool used in the Pacific Region. SCOPIC was initially developed during the 8-year Pacific Island Climate Prediction Project (PICPP) in order to provide Pacific Island Countries (PICs) with a flexible, standalone PC probabilistic prediction tool. Although the primary use of SCOPIC is to provide rainfall outlooks, it can produce predictions for a range of temporal data provided there is sufficient information with which to train the statistical model. Furthermore, SCOPIC contains a comprehensive drought analysis module.

SCOPIC version 3 has recently been released to partner PICs within the Climate and Ocean Support Program in the Pacific (COSPPac). There has been significant interest in the use of Predictive Ocean Atmosphere Model for Australia (POAMA) forecasts for the Pacific Region due to the recent conversion of the Australian Bureau of Meteorology Seasonal Climate Outlook from the statistical model to the POAMA dynamical model.

This paper has examined the skill of the seasonal outlooks produced by both SCOPIC and POAMA over the February 2012 to February 2013 period for selected COSPPac partner Pacific Island Country (PIC) stations. It must be stressed that this is too short a period for any conclusions about the long-term relative performance of the two forecast systems to be drawn.

This skill comparison analysed Linear Error in Probability Space (LEPS) scores and hit rates for each station. Statistical analysis has shown that: Kiribati had the highest POAMA tercile hit rate and the highest POAMA LEPS skill; the Cook Islands had the highest SCOPIC hit rate; and Vanuatu had the highest SCOPIC LEPS skill. Overall for the periods starting February 2012 to February to 2013, POAMA had higher skill in both LEPS and hit rate for seasonal climate outlooks for a subset of Pacific Island Countries. However, POAMA is currently limited by the number of available station outlooks for the Pacific Region due to downscaling issues; further analysis will be possible when additional outlooks are made available. Only selected PIC stations were examined due to this POAMA limitation, and as PICs can experience significant in-country topographic and climatological variation between meteorological measurement stations; SCOPIC remains a more useable tool to model rainfall variation in the Pacific Region.