Optimal crop planning using a chance constrained linear programming model

Abstract

A monthly time-stepped chance constrained linear programming (CCLP) model was developed to derive optimal cropping patterns and optimal operational strategies for the Sri Ram Sagar Project, with stochastic inflows. The stochastic nature of the inflows was incorporated into the model in its equivalent deterministic form. These equivalent deterministic inflow values were estimated from the annual and monthly probability distribution of observed inflows, and named the chance constrained linear programming model-annual and chance constrained linear programming model-monthly, respectively. The models were solved for nine different dependable inflow levels, namely for 50, 55, 60, 65, 70, 75, 80, 85 and 90% in each CCLP. The results of the models were compared with respect to net benefit, irrigation intensity, total cropped area, optimal cropping pattern, optimal releases, evaporation loss and initial storages. Based on the results obtained, it can be concluded that, in CCLP optimization, the probability distribution of ‘model time period’ (t-month in this case) should be considered rather than the probability of ‘planning time period’ (year in this case).