IS It POSSIBLE TO PREDICTIVELY MODEL SHALLOW KARST AQUIFERS AT A MEANINGFUL SCALE?

The Savoy Experimental Watershed (SEW) has been the focus of about 30 theses, dissertations, and journal articles which describe a body of research that encompasses karst-aquifer response across a wide range of hydrologic conditions that have occurred since 1996. These studies indicate that more than ten key variables play a major role in controlling ground-water flow and add significant three-dimensional complexity to flow paths at scales ranging from tens to thousands of meters. Utilizing data from surface geophysics, ground-penetrating radar, drilling, borehole geophysics, surface-water/ground-water interaction, base-flow recession of springs, continuously monitored water levels in wells, multiple tracer tests run at the same sites under different hydrologic conditions, and ground-water temperature and specific conductance determinations at SEW, we document temporal and areal variability far beyond our original conceptualization of the system. Each field season brings new insight from an expanding range of empirical sources, and each storm points out new facts we did not previously know. Our understanding is being continually revised. We document spring basin boundaries changing with fluctuating water level, flow directions shifting, and flow paths being plugged and other flow paths washed clean from storms of nearly identical duration and intensity. Although we have gained significant understanding by using models to test our hypotheses about the SEW karst system, the implications from our data sets is that predictive modeling of karst aquifers at site-specific to basinal scales will most likely be incorrect and misleading, and as a result, will be counterproductive to conveying meaningful science to the public.