EVALUATION OF A SIMPLE TECHNIQUE FOR ESTIMATING TWO-DOMAIN TRANSPORT PARAMETERS

Abstract

A simple method has been proposed for estimating the fraction of mobile water (β) and the solute transfer rate coefficient (ω) of the mobile-immobile (MIM) solute transport model. The method is based on a log-linear solution to the MIM equations that are simplified by assuming that dispersion, D, is negligible and that the tracer concentration in the mobile domain is equal to the input tracer concentration. We evaluated these assumptions by computing parameter estimates with the log-linear solution and the solution to the complete MIM equations with D preset to a negligible value using synthetic data sets created for a range of transport conditions. We found the assumption of negligible D to be valid for Peclet numbers, P, > 100 at any ω and for ω < 0.01 at any P. For P < 0.3, D is too large to assume negligible and obtain accurate parameter estimates for most ω values. At ω > 1, the solute transport system effectively degenerates into a single mobile domain, and the approach performs poorly for most P values. At intermediate P and ω conditions, the accuracy of the method depends on the specific transport conditions determined by P, ω, and β. The log-linear solution to the MIM equations provided worse estimates of β and ω across much of the ω and P range reported in the literature than the complete solution with a small preset value of D. A modified version of the log-linear solution that relaxed the assumption of resident tracer concentration equal to input concentration yielded accurate parameter estimates for a wider range of conditions than the log-linear solution but a smaller range than the complete solution. When applied to a set of measured data from two soils, the three estimation methods gave similar values for β and ω values within a factor of 1.5.