Methodological Problems in Simulation of the Agro-Ecosystems

Abstract

Simulation models describe the functioning of agro-ecosystems and the flow of processes in them using computer programs. With their help can be made computer experiments and to predict the behavior of agro-ecosystems. One of the main problems in the creation of simulation models is imitation of uncontrollable soil and weather factors. A number of studies of soil properties and meteorological characteristics indicate that these stochastic variables do not follow the normal (Gaussian) distribution. In the selection of statistical models that approximate the distribution of such stochastic variables, the most appropriate is a family of Johnson distributions. The article examines ways to generate pseudorandom numbers from Johnson and Burr distributions families for simulating soil and weather factors determining agro-ecosystems.