Operational forecasting of South African sugarcane production: Part 1 – System description

Commercial sugarcane crops in South Africa are grown under a wide range of agronomic and socio-economic conditions. These factors, together with climatic variation have resulted in a 17% variation in sugarcane production and there is considerable scope to improve productivity through accurate and timeous forecasts. This paper reports on the development of an operational crop forecasting system based on a simulation model. The country’s entire area of sugar production was subdivided into homogeneous climate zones using a wide range of data and expert opinion. These zones serve as simulation units within the system and model input and area aggregation data were obtained for each climate zone. Irrigation is simulated according to typical, zone specific strategies taking into account water use restrictions. Simulations of crops growing in the current year are completed using 10 historic seasons to substitute the remainder of the season. The selection of these seasons is based on the climate outlook. Reports containing information for national, regional and site specific cane production are generated and distributed to industry stakeholders. To the authors’ knowledge, this is the first national scale model-based operational yield forecasting system for sugarcane. Possible future improvements to the system may include stochastic input variables, more representative irrigation simulations, quantifying forecast uncertainty and providing suitable reference crop yield. The system is evaluated in another paper.