Exploring solute transport and streamline connectivity using lidar-based outcrop images and geostatistical representations of heterogeneity

Abstract

Using high-resolution lidar scans of a braided stream deposit, we investigate solute transport characteristics and streamline-based connectivity that are lost when simulating the outcrop heterogeneity using geostatistical methods based on two-point covariance functions. Attributes of the lidar scans were used to segment the outcrop into sand- and gravel-dominated lithofacies. Simulated fields were created using sequential indicator methods based on the two-point covariance of the binary segmented lidar field. Sand and gravel lithofacies are then assigned reasonable hydraulic conductivity values. Two-dimensional advective-diffusive solute transport simulations in the segmented lidar field show strong solute focusing through gravel-dominated strata, resulting in a heavy-tailed (e.g., non-Fickian) breakthrough. The sequential indicator fields do not replicate the early and late time arrival characteristics. Streamline-based analysis shows that the sequential indicator fields do not reproduce connectivity of the segmented lidar field. Even when the sequential indicator fields are highly conditioned, streamlines migrate between sands and gravel beds nearly twice as often as streamlines in the segmented lidar field. Copyright 2009 by the American Geophysical Union.