Risk assessment and food security

Risk and uncertainty affect all decisions and result in inefficiencies in the agricultural sector as well as insecurity in food supply in many countries. Carefully validated crop models have a role to play in addressing these issues, because they can be used to isolate and quantify risk associated with weather variability and economic risk arising from uncertainty in the costs and prices of production. The methodologies for risk assessment in the DSSAT software are reviewed and illustrated. These are based on Bernoullian utility theory and stochastic efficiency criteria. At the household level, the models have been used to derive site-and season-specific crop management practices that reduce risk in southern Africa. At the regional level, a pilot information system has been constructed that links crop models, satellite-derived estimates of rainfall in real time, and a geographic information system, to obtain millet yield forecasts for districts in Burkina Faso that can be updated regularly through the growing season, to give early warning of impending pro­duction shortfalls. While work is still required to improve the methodologies used, the potential of models for providing information that can help reduce down-side risk, and thereby improve food security at the household and regional levels, is considerable.