A Dynamic Risk Optimization Model for Evaluating Profitable and Feasible Water Management Plans

Abstract

ABSTRACT Currently the South African government is advocating the cultivation of high-valued crops and more efficient use of available water resources through the adoption of more efficient irrigation technology and irrigation scheduling. A requirement of the National Water Act (Act 36 of 1998) is the compilation of water management plans. The main objective of this paper is to develop a multiperiod mathematical risk programming model able to assist water user associations with the compilation of water management plans that are both profitable and feasible. Special care was taken to represent canal capacities and irrigation system application rates in the model. Risk simulation procedures are used to generate an appropriately correlated inter- and intra-temporal risk matrix for the programming model. A combination of subjectively elicited distributions of crop yield and objective data on crop prices were used to characterise risk. The model was applied to a representative flood irrigation farm in the Vaalharts irrigation scheme South Africa to demonstrate the capability of the model to optimise agricultural water usage over a 15 year planning horizon. Model results clearly indicate the potential of high-value crops and more efficient irrigation technology to reduce the impact of water restrictions. Furthermore infrastructure, the financial position of the farmer and the level of risk averseness have significantly impacted on the results. Policy makers and government authorities should take cognisance of these factors when evaluating water use efficiency and water management plans of different water user associations. Improvements to the adop ted modelling procedure are also suggested. JEL classification: C6, Q15, Q12 Keywords: Dynamic Linear Programming, risk, irrigation, feasibility, South Africa