Field study of dispersion in a heterogeneous aquifer: 2. Spatial moments analysis

TLDR

The study builds on the stochastic theory of Gelhar and Axness (1983) to analyze dispersion in a highly heterogeneous aquifer based on spatial variability of hydraulic conductivity. The authors analyzed a 20‑month natural‑gradient tracer experiment, performed sensitivity studies on mass loss, and interpreted plume moments using two simple models—pure advection and advection plus dispersion in a converging nonuniform flow field. The tracer data exhibit strongly non‑Gaussian concentration distributions, systematic mass loss, and a longitudinal dispersivity of 5–10 m, which is higher than values from other heterogeneous sites but lower than estimates assuming uniform flow.

Abstract

Analysis is performed of a 20‐month natural gradient tracer study in the saturated zone of a highly heterogeneous aquifer. Graphical presentation of concentration distributions versus time and spatial moments analysis reveal dramatically non‐Gaussian behavior and a systematic mass loss. Implications of the mass loss on plume moments is analyzed through sensitivity studies. The moments data are interpreted by applying two simple models: (1) pure advection from a continuous source, and (2) advection plus dispersion in a converging nonuniform flow field. A longitudinal dispersivity of 5–10 m is estimated from the latter model and is somewhat larger than the value of about 1.5 m calculated by Rehfeldt et al. (this issue) using the stochastic theory of Gelhar and Axness (1983) based on independent measurements of the spatial variation of hydraulic conductivity. The dispersivity of 5–10 m is an order of magnitude larger than values measured at recently studied field sites (Borden and Cape Cod) with less heterogeneity, but an order of magnitude lower than would be computed from the moments data if the flow is presumed to be uniform.