Evaluation of empirical equations and intelligent model (ANN and ANFIS) for prediction of Dispersion coefficient in river

Water pollution in rivers is a complex process. Transmission of pollutant in river is one of the most important problems in Environmental Science. The governing Equation of transmission of pollutant in river is Advection Dispersion Equation (ADE). The ADE is Partial differential equation (PDE) that it’s one of the most important equation that used in fluid engineering Special water engineering، this equation named in general Convection equation. This equation has two part، Advection and dispersion term. To modeling of pollution transport in rivers needed to developing Numerical model for solving the ADE and calculating of Longitudinal Dispersion coefficient (LDC). LDC is one the important parameters in the Pollution transport simulations. Accurate estimates of this parameter increase the accuracy of the computer modeling. There are many methods to calculate the diffusion coefficient of the experimental and mathematical methods. The results of the experimental has significant errors because it’s based on indirect parameter in the simplification of modeling. So it needs to be calibrated with observational data or Methods that the Dispersion coefficient can be calculated directly. In this study، the routing method was used to calculate the Dispersion coefficient and the following a careful review was done on the empirical formulas. Final result shows that the intelligent method (ANN and ANFIS) is more accurate in compare with empirical formulas.