2015 GSA Annual Meeting in Baltimore, Maryland, USA (1-4 November 2015)

Paper No. 4-10
Presentation Time: 10:30 AM

MAPPING SUBSURFACE HETEROGENEITY TO ESTIMATE TRANSPORT CHARACTER OF THE BULK FUELS SPILL SATURATED PLUME, KIRTLAND AIR FORCE BASE, ALBUQUERQUE, NEW MEXICO


GATZ-MILLER, Hannah, Department of Earth and Planetary Sciences, University of New Mexico, MSC03-2040, 1 University of New Mexico, Albuquerue, NM 87131 and WEISSMANN, Gary S., Earth and Planetary Sciences, University of New Mexico, MSC03-2040, 1 University of New Mexico, Albuquerque, NM 87131-0001, hgatzmiller@unm.edu

In 1999, Kirtland Air Force Base (KAFB) in Albuquerque, New Mexico, reported leakage of an unknown amount of jet fuel from an underground pipe. Subsequent investigations revealed that the jet fuel had leached through the vadose zone to contaminate the Santa Fe Group Aquifer, which the city of Albuquerque relies on for its municipal water supply. There are two main phases of the saturated plume: a solute phase, and a light, non-aqueous phase at the top of the saturated zone (smeared by a rising water table). Monitoring wells have indicated that the solute phase plume is migrating north towards the cone of depression created by the Albuquerque municipal wells. We are developing a subsurface model to characterize this solute phase plume migration.

To characterize the heterogeneity of the system, lithologic logs, geophysical logs, and core data from KAFB boreholes were collected and analyzed with transition probability geostatistics (TPROGS). Previous studies of the Santa Fe Group, corroborated with core data from KAFB boreholes, indicates that the material at the saturated zone consists mainly of ancestral Rio Grande River deposits, with grain sizes ranging from silty sands to large, subrounded gravels, and channel deposits oriented generally north-south. Mean lengths of facies used to develop the Markov chain model were estimated from core and well log data in the vertical direction and horizontal bar and channel lengths measured from modern analogs, including the modern Rio Grande River through use of historical (1949 and 1961) aerial photographs, and other braided stream systems described in the literature. We varied the mean lengths across a reasonable range in different Markov chain models in order to test the sensitivity of models to these length scales. Incorporation of these models of spatial variability will ultimately result in a more accurate understanding of the direction and rate of movement of the contaminated groundwater plume.