Effect of distance-dependent dispersivity on density-driven flow in porous media

In this study, the effect of distance-dependent dispersion coefficients on density-driven flow is investigated. The linear asymptotic model, which assumes that dispersivities increase linearly with distance from the source of contamination and reach asymptotic values at a large asymptotic distance, is employed. An in-house numerical model is adapted to handle distance-dependent dispersion. The effect of asymptotic-dispersion on aquifer contamination is analyzed for two tests: (i) a seawater intrusion problem in a coastal aquifer and (ii) a leachate transport problem from a surface deposit site. Global Sensitivity Analysis (GSA) combined with the Polynomial Chaos Expansion (PCE) surrogate modeling is conducted to assess the influence of the dispersion coefficients on the contamination plume for both configurations. For the seawater intrusion problem, the results show that the length of the toe is mainly controlled by the asymptotic transverse dispersivity whereas the spread of the concentration is sensitive to the asymptotic longitudinal dispersivity and the asymptotic dispersivity distance. The latter is the most important parameter controlling the amount of salt which intrudes into the aquifer. For the leachate transport problem, the results show that the asymptotic longitudinal dispersivity coefficient does not affect the concentration distribution. The asymptotic dispersivity distance has a strong effect on the total amount of contaminant that enters the aquifer. This effect can be three times more important than the effect of the asymptotic transverse dispersivity. These findings are likely to be helpful for the investigation and management of density-driven flow problems.