GEOSTATISTICAL AND DETERMINISTIC MODELING OF ALLUVIAL FAN HYDROSTRATIGRAPHY AT THE LAWRENCE LIVERMORE NATIONAL LABORATORY

We developed 3D stratigraphic models to provide a basis for groundwater flow and contaminant transport simulation in order to test remediation strategy effectiveness at the Lawrence Livermore National Laboratory. Analysis of core shows that it is possible to construct a stratigraphic framework using paleosols to define individual stratigraphic units, related to previously defined hydrostratigraphic units, to aid in the correlation of channel and overbank deposits of the complex alluvial fan system. Geophysical logs from 19 wells, from which 15 have had continuous core recovered, provide data for this study. Depths of the wells range from 44.2 to 64.0 meters and were drilled on an approximately 7.6 meter grid. Four sediment types dominate the alluvial deposits – sandy clay-rich gravel and cobbles from channel deposits, fine to medium sand from channel deposits, sandy silt from overbank deposits, and relatively-mature paleosols that developed as a result of extended periods of fan exposure. Using data collected from core, geophysical logs, and hydraulic tests, two stratigraphic models were produced -- one using transition probability geostatistics and the other using deterministic methods based on surface generation using the TOPOGRID function in ArcInfo. These models honor channel orientations and distributions observed from well data and promise to be extremely useful in modeling groundwater flow, contaminate transport, and remediation strategies.