Delimiting and quantifying seasonal changes in hyporheic extent using inverted electrical resistivity data

Abstract:

As stream water flows, some of the surface water moves into and back out of the streambed where it also mixes with groundwater. This area of mixing, the hyporheic zone (HZ), has been identified as a critical location in which streams process nutrients and organic matter. Consequently, efforts to model or predict water quality in streams usually require that the extent and exchange rates of the HZ be estimated. Yet, there are no widely accepted methods that can directly and objectively characterize the spatial distribution of this mixing and how it changes over time. Time-lapse electrical resistivity (ER) imaging has been used to obtain 2-D representations of HZ extent and exchange dynamics during tracer studies; however, it is impossible to exactly estimate area from 2-D models of the 3-D subsurface. Past studies have utilized inverted ER data to examine relative areas through time, but interpretation of effective hyporheic extent often relies on selection of arbitrary signal thresholds, to which areal estimates are sensitive. To address this challenge, we introduce a novel method that applies time-series clustering algorithms to ER data to delineate both the extent of the HZ as well portions within the HZ with similar patterns of connectivity to surface water. We present results using this new workflow to assess changes in HZ connectivity in mountain streams throughout seasonal flow recession. Our approach can inform data-driven efforts to assess spatial variability of hydrologic processes at single time points as well as to determine causes of temporal variability in biogeochemical phenomena.