Cordilleran Section - 98th Annual Meeting (May 13–15, 2002)

Paper No. 0
Presentation Time: 2:45 PM

COMPARISON OF METHODS TO ESTIMATE SEEPAGE BETWEEN STREAMS AND GROUND WATER, WILLAMETTE VALLEY, OREGON


CONLON, Terrence D. and LEE, Karl K., U.S. Geol Survey, 10615 SE Cherry Blossom Drive, Portland, OR 97216, tdconlon@usgs.gov

Understanding the connection between surface water and ground water is important for managing both resources. Quantifying the hydraulic properties of the streambed and the magnitude of seepage between the stream and ground water system is critical to this understanding. Traditionally, seepage is calculated from seepage runs, in which the seepage into or out of a stream is the difference between the upstream and downstream discharge measurements, after accounting for diversions and tributary inflows between the measurement locations. In the Willamette Valley, Oregon, where streams flow on fine-grained, low permeability material, and where quantifying streamflow diversions during the irrigation season is difficult, the calculated seepage is often less than the uncertainty of the measurements. To provide a comparison with, and possibly an alternative to, seepage runs, seepage was estimated stream reaches in the Willamette Valley by three other methods: baseflow separation, seepage meters, and temperature gradient method. Seepage runs and baseflow separation integrate seepage over a stream reach but also include uncertainties of diversions and inflow along the reach, while seepage meter and the temperature gradient methods provide point measurements and do not include uncertainties associated with quantifying diversions and inflow. The temperature gradient method also provides head gradient information to identify conclusively the direction of seepage and an estimate of the permeability of the streambed. Preliminary results from this study indicate that in the Willamette Valley, seepage estimated by the three methods ranges from 0.5 x 10-8 to 88 x 10-8 m/s. Generally, estimates by seepage meter are lowest and estimates by temperature gradient and baseflow method are highest. Differences in scale of measurement and heterogeneities in head gradient and streambed permeabilities probably account for much of the variation. These methods provide a range of possible streambed seepage rates with which to calibrate flow models and identify the direction and magnitude of seepage.