CR 30:239-246 (2006)  -  doi:10.3354/cr030239

ABSTRACT: Sugarcane production is an economically important activity in South Africa. Climate variability causes high fluctuations in annual production, and stakeholders in the industry are often compelled to make crop management, harvesting, transport, milling and marketing decisions under high levels of uncertainty. In this study, a previously derived model-based yield forecasting system is used to produce hindcasts of sugarcane production in South Africa. The derivation of climate outlooks by the South African Weather Service (SAWS) and a technique to translate these into yield forecasts are explained. Model generated forecasts are consistently more accurate than forecasts based on climatology and, depending on the time of the year, capture between 11 and 58% of the natural variation in national sugarcane production. Climate outlook information generally enhances production forecasts, suggesting that these outlooks are sufficiently accurate for decision making and that the translation techniques perform satisfactorily. Different regions, however, display different accuracy responses to the climate outlook and several suggestions for future research are made, such as incorporating different climate forecasts and creating consortiums between climate forecasters, crop modellers and decision makers.

KEY WORDS: Seasonal climate outlook · Production forecasting · Sugarcane · Climate forecast application