MODELING SEDIMENTOLOGY AND AQUIFER HETEROGENEITY USING LIDAR AND DIGITAL PHOTOGRAPHS OF OUTCROP ANALOGS

In order to evaluate the 3D distributions of lithofacies and model the influence of heterogeneity on subsurface fluid and contaminant movement, we acquired digital imagery and lidar scans from four different outcrop exposures that encompass a range of fluvial and glacio-lacustrine depositional processes and scales. In our first study, we excavated a series of exposure panels in a sand quarry through lacustrine delta sediments of the Borden aquifer, mapped and measured lithofacies positions in these exposures, made facies-based measures of hydrogeologic properties, used image processing techniques to segment and classify images for lithofacies mapping, and modeled the three-dimensional facies distributions using transition probability geostatistics. The resulting realizations are used in numerical experiments that explore the impacts of geologically grounded sedimentary aquifer property heterogeneity on contaminant transport. At a second site, we mapped amalgamation and distributions of channel belts at an exposure of a distal Salt Wash Member (Morrison Formation). These results will aid evaluation of interconnectivity of high permeability facies and evaluation of 3D geometries of facies in highly complex distal fluvial settings. Our third site focused on characterization and modeling of a generally coarsening upward fluvial succession in Westwater Canyon Member (Morrison Formation) across an approximately 300x400-m amphitheater exposure. In this case, we are evaluating semi-automated methods for segmenting and classifying 3D facies distributions to create realistic training images for input into a multiple point geostatistical simulation. At a fourth site, we collected lidar scans from a large (1-km) amphitheater canyon in the Paleocene Nacimiento Formation to evaluate cyclicity and facies connectivity in this fluvial succession. In each case, we are able to construct outcrop surface models of the site, and then interpret lithofacies from imagery in order to create virtual outcrop models that can be used for 3D modeling. Outputs from these models can then be incorporated into fluid flow models in order to evaluate the influence of physical heterogeneity on fluid movement and contaminant transport.