Use of the Southern Oscillation to predict Australian sorghum yield

Abstract

Grain sorghum is an important, rain-fed, Australian summer crop, and previous work has demonstrated that sorghum yield is moisture-limited. Rainfall in and before the growing season is related to the Southern Oscillation, a weather phenomenon which possesses a strong “memory” and thus the potential for long-range prediction. Sorghum yield anomalies have been correlated with values of an index of the Southern Oscillation from during and before the growing season. Significant negative correlations were found, even with values of the index from well before the time of planting. This indicates that early prediction of sorghum yield is feasible through the use of the Southern Oscillation index. Even stronger correlations were found between sorghum yield and trends in the Southern Oscillation index in the months leading up to the planting season. Predictions made from the trend in the index up to a month prior to the start of the planting season account for about 50% of the variance in yield, after removal of long-term yield trends resulting from changes in production technology. The strength of the relationship, and the early availability of the forecasts, appear to offer a potentially useful long-range crop-forecast system.