Abstract
The El Niño Southern Oscillation (ENSO) template concept is presented for seasonal streamflow prediction methodology by considering the Southern Oscillation Index (SOI) and Sea Surface Temperature (SST) variables. The methodology provides linguistic and quantitative inference capabilities. The prediction model uses the geostatistical (Kriging) technique. Each ENSO template has nine categories including one with high SST and low SOI values that represents the El Niño event. Similarly, the category with low SST and high SOI values depicts the La Niña event. The application of the methodology is presented for the seasonal streamflow records in the southeastern part of the Australian continent along the Pacific Ocean. April-September streamflow is predicted by using five different lead times including 3-month ENSO indicator averages. The seasonal streamflow predictions at different lag times are obtained given the values of SST and SOI. The overall relative prediction error is rather small at about 13%. The bigger the lag the bigger is the prediction error. However, the relative error between averages of observation and prediction values is less than 5%. Similar ENSO templates can be used for streamflow prediction in other parts of the world.