2002 Denver Annual Meeting (October 27-30, 2002)

Paper No. 11
Presentation Time: 4:30 PM


ROHE, Michael J.1, SMITH, Richard P.2 and MCLING, Travis2, (1)Applied Geosciences Department, Idaho National Engineering and Environmental Lab, P.O. Box 1625, Mail Stop 2107, Idaho Falls, ID 83415-2107, (2)Geosciences Research Department, Idaho National Engineering and Environmental Lab, P.O. Box 1625, Mail Stop 2107, Idaho Falls, ID 83415-2107, mil@inel.gov

The Eastern Snake River Plain aquifer beneath the Idaho National Engineering and Environmental Laboratory (INEEL) is a highly heterogeneous system comprised of sequences of sediment and basalt flows that range from fractured to massive. This aquifer is monitored for water level changes in several hundred INEEL wells. Hydrographs of nearly 100 aquifer wells were examined to determine correlations between regional recharge and water level fluctuations.

Many of these well hydrographs extend for 50 years and show long-term oscillations and an overall decreasing trend. The trend of decreasing water level may be the result of increased groundwater use or regional climatic trends. The long-term oscillations are of a floating frequency with a period ranging from 10-15 years and appear to be climate-related. Snowpack surveys from surrounding mountain ranges and streamflow records from tributary valleys show low-frequency cyclic behavior of similar period. We are removing seasonal and trend effects from the snowpack, streamflow, and aquifer water level data, using the Kolmogorov-Zurbenko (KZ) filter, to provide stationary data for the analysis of spatial and temporal distribution of time-series correlations. The KZ filter, based on iterative moving averages, can remove signals of selected frequencies by modifying the filter window length and number of iterations.

Calculations of aquifer hydraulic gradient, determined from the plane surface formed by the water levels in three wells, have resulted in another time-series. The near-decadal cycle is also seen in the aquifer hydraulic gradient. The seasonal component results in ten percent or less of the total variation in hydraulic gradient whereas some sixty-percent or more of the total gradient variation is attributable to the long-term, near-decadal oscillation. Historically, modeled contaminant plume configurations, based on an assumed constant aquifer gradient, do not match configurations observed in the aquifer beneath the INEEL. The variation in aquifer hydraulic gradient, resulting from near-decadal climate changes, may explain these discrepancies.