“INCOMPLETE SYMPHONY”—LACK OF HIGH-FREQUENCY RECORDS FAILS TO CAPTURE THE COMPLEXITY OF KARST AQUIFERS

Kirchner et al. (2004) drew an analogy between music and hydrology, paraphrased below:

“Imagine trying to understand a Beethoven Symphony if we could only hear one note every minute or two! This is what we are trying to do when we infer the hydrological functioning of a catchment from weekly or monthly grab samples. Or imagine trying to understand a symphony from a high-fidelity recording of just one of its crashing crescendos. That is what we are trying to do when we analyze high-frequency samples of an individual storm event. Continuous high-frequency monitoring of catchment hydrochemistry will require significant resources and tenacity. In our view, however, what we stand to learn is well worth the effort. If we want to understand the full symphony of catchment behavior, then we need to be able to hear every note.”

We feel this analogy is particularly apt for karst. Hydrogeologic datasets from the Savoy Experimental Watershed (SEW) in northwest Arkansas provide examples of high-frequency sampling and illustrate the benefits of this sampling rationale to gain enhanced understanding of a mantled-karst aquifer. More than 25 years of research at the SEW has garnered numerous sampling events from interdisciplinary collaboration. Coupled with continuous records, these data have provided conceptual constrains to a model of complex flow and processes that would not be available using “grab-sample” collection methods alone.

Drawing on near-continuous records of precipitation, temperature, discharge, and tracer tests of SEW springs, we have been able to craft a conceptual model that explains all empirical observations. The spring system is interpreted as being a subsurface lake, dammed by faulting and sediment infilling on the margin of a graben. Primary groundwater conduits follow orthogonal joints that have been enlarged by dissolution that drain to 2 major springs. Multiple input sources provide recharge upgradient from the fault. This system allows groundwater velocities of greater than 700 m/d, with base-level springs that resurge as artesian flow below the overlying continuous chert layers in the Boone Formation. The efforts and costs associated with high-frequency sampling can be significantly greater than isolated sampling and data collection, yet the insight gained from “hearing every note” to us is invaluable.