Derivation of Optimal Equations for Soil Erosion in Furrow Irrigation UsingMulti-Gene Genetic Programming Based on Pareto Optimal Solution(POMGGP)

Soil loss is one of most important parameters in design and operation of furrow irrigation systems. Design of furrow irrigation systems is often done based on hydraulic concepts. Because of the loss of predictive equations for soil loss, fertilizer runoff, these losses cannot be estimated quantitatively. On the other hand its measuring is expensive, time consuming and difficult. The main aim of this study is to using the ability of intolerance calculation in derivation of soil loss equations for furrow irrigation by using multi Genetic Programming based on Pareto Optimal solution (POMGGP). So, a database of field measurements of soil erosion in furrow irrigation is collected and multi Genetic Programming in combined with Pareto Optimal solution and preprocessing of maximum dissimilarity for train and test sets is used. Fourteen different models for soil loss are developed and the best equation is derived. The results based on statistical indices of R, RMSE, MAE, RAE and NSE has shown that the POMGGP has the ability in finding optimum equation and parameters for furrow irrigation soil loss estimation and can be used in design and operation of furrow irrigation systems.