Optimization models in agriculture and natural resources: the validation and verification process

The optimization tools have been extensively used for developing the agriculture production planning decision support tools all over the world during several past decades. Although many different models based on different assumptions and mathematical methods have been developed, very few deal with the verification and validation procedures for these optimization models. This stems mainly from the fact that each such optimization problem is unique, hence no general validation and verification process can be designed. This paper aims to show the possibilities of developing a verification and validation process for the particular problem of optimizing the sowing plan when the randomness of harvests and the crop succession requirements are considered. A stochastic programming model is applied, the verification of the model and the resowing constraints are discussed and the validation procedure based on Monte Carlo simulations is suggested and performed for the particular case of a South Moravian farm.