Analysis of Virus Transport in Groundwater and Identification of Transport Parameters

Abstract

The present study deals with the analysis of virus transport in groundwater and estimation of transport parameters. The parameter estimation is formulated as a least-squares minimization problem in which the parameters are estimated by minimizing the deviations between the model predicted and experimentally observed virus concentrations. A parameter estimation procedure is developed by coupling a hybrid finite volume numerical model simulating one-dimensional virus transport with the Levenberg-Marquadart algorithm. The numerical model employs a globally second order accurate explicit finite volume method for the advective transport and an implicit finite difference method for the dispersive transport. The comparison of model prediction with analytical solution shows that the numerical model accurately predicts virus movement in both advective and dispersive dominated transport. The parameter estimation indicates that for the case of estimating more than three unknown parameters, the inverse procedure results in nonunique optimal estimates. Further, for the case of estimating two or three unknown parameters, the presence of inactivation coefficients of liquid and sorbed phases also results in nonunique estimates. It is concluded that a priori estimation of one of the inactivation coefficient is necessary for unique estimation of other unknown parameters. A detailed statistical analysis is carried out to study the effect of bias induced by the objective function on the estimated parameters when the data contains the errors. It is found that in the presence of noise in the virus concentration data, the objective function does not induce any bias on the parameter estimates as the true value falls within 95% confidence levels at all noise levels.