Master Diagnostic Curve for Dispersion Coefficient of Soils

Abstract

A master diagnostic curve (MDC) method is proposed for identifying the dispersion coefficient from observed breakthrough curve in soil–column experiments. The method uses matching of diagnostically plotted points to the MDC. Accurate identification of the dispersion coefficient is possible with a parallel shift of only one axis. Another advantage of the method is that the MDC is an invariant curve and once prepared, can be used for different data sets characterized by high Péclet number. The new method is simple, does not require large computations as in conventional least-squares approach, and yields a quick and accurate estimate of the dispersion coefficient. The new method does not require an initial guess for the dispersion coefficient. It can also yield a value of the dispersion coefficient from a single observed concentration. Although subjectivity is involved in matching, there is a transparent visual realization of the reliability of the estimated dispersion coefficient, and the points with substantial errors. The new method has several advantages over the least-squares approach and is suited for an advanced study on the subject. An estimate of the dispersion coefficient obtained using the new method is as good as obtained using a good optimization method.