SCOPIC: Seasonal Climate Outlooks in Pacific Island Countries

What is SCOPIC?

SCOPIC is a decision support system for generating probabilistic predictions (seasonal climate outlooks) for rainfall, temperature or other climate related parameters.

Why was it developed?

SCOPIC was developed to provide Pacific Island nations with a standalone PC version of the Commonwealth Bureau of Meteorology's operational seasonal climate prediction system. Based on historical data, the system uses a statistical method called linear discriminant analysis. Forecast probabilities of the variable we're interested in (called the predictand), e.g. rainfall, are generated via historical relationships with different predictors.

What can it predict?

SCOPIC can predict whatever you want it to, provided you have enough historical data to train it! The predictand data will in most cases consist of rainfall or temperature, although in principle, any data could be used. For instance, the predictand could consist of tonnes of taro, or water storage volume. The predictand needs to be in monthly format, saved in a Microsoft EXCEL file and consist of at least 25 years of data. SCOPIC can display forecasts on regional maps.

What predictors does it use?

SCOPIC is very flexible allowing literally any type of monthly time-series data to be used as a predictor in generating the forecast. However, discretion must be used when trying different combinations of predictor and predictand in the hope of finding something that 'works'. There is a very real risk of introducing 'artificial skill' by such a practice. Two predictors are supplied with the installation package of SCOPIC: sea surface temperature (SST) anomalies and the Southern Oscillation Index (SOI).

The Bureau's system uses indices to represent the strengths and phases of broad patterns of SST anomalies over the Indian and Pacific Oceans. Complex spatial patterns are reduced to individual numbers (Empirical Orthogonal Functions) via Principal Component analysis of SST anomalies over the two ocean basins. Regions of unusually cool or warm ocean water are indicative of, or can promote persistent atmospheric anomalies, especially at crucial times of the annual cycle. They are useful predictors because they change relatively slowly. The primary predictor is SST1 or EOF1, which represents the El Niño-Southern Oscillation phenomenon.

So what does SCOPIC tell me?

SCOPIC generates forecasts in the form of probabilities (chance, likelihood) of being in either two or three categories which are, by definition, equally likely over the historical record. The two-category system provides chances of being above or below the median (historical chance of 50% or ½), while the three-category system provides chances of being in tercile ranges (historical chance of 33.3% or ⅓). Accompanying these are graphical and text outputs from supporting analyses. These include skill tests, hindcasts, data-browsing, statistical analyses and scatter plots.