EXPECT THE UNEXPECTED WHEN QUANTIFYING EXCHANGE BETWEEN GROUNDWATER AND SURFACE WATER (Invited Presentation)

Fluxes of water and chemicals across the sediment-water interface are conceptually simple but surprisingly difficult to quantify. Problems can be divided into three categories: conceptualization, complexity, and confusion. Although we know flow lines are curvilinear at and near the shoreline, we commonly assume that flow is virtually horizontal adjacent to a surface-water body and vertical beneath it, assumptions that are often incorrect. We also know that rates of exchange should decrease exponentially with distance from the shoreline, but after decades of measuring seepage there still is no consensus regarding seepage distribution or how to apply it. Sometimes seepage decreases with distance from shore; sometimes it does not. Part of the reason for data not matching theory is due to the large degree of complexity and heterogeneity in this dynamic setting. Variability in shoreline location, bed topography, sediment character, and water chemistry should be expected in fluvial settings, but this variability also occurs in lake and wetland settings commonly assumed to be much more temporally stable. Catastrophism also occurs in these settings at multiple scales, all of which can exert a strong influence on magnitude and distribution of seepage fluxes. Confusion comes into play when we do not understand all of the processes that collectively control seepage. Variable salinity and water density can have a profound effect on seepage distribution and rate. Sediment compressibility, sometimes associated with accumulation, compressibility, and release of trapped gas, can confuse interpretation of seepage data. Biology can also be an important factor; many animals living on or within the sediment can create flow at rates that greatly exceed gradient-driven seepage. When combined with natural and anthropogenic temporal variability due to rain, wind, waves, evapotranspiration, and even boat traffic, variability that rarely is measured with current technology, it is no wonder that data often are surprising and very difficult to interpret. New tools and better use of existing methods will help to address these issues and improve our collective understanding at this dynamic interface.