Interactions among hydraulic conductivity distributions, subsurface topography, and transport thresholds revealed by a multitracer hillslope irrigation experiment

Abstract

Interactions among hydraulic conductivity distributions, subsurface topography, and lateral flow are poorly understood. We applied 407 mm of water and a suite of tracers over 51 h to a 12 by 16.5 m forested hillslope segment to determine interflow thresholds, preferential pathway pore velocities, largescale conductivities, the time series of event water fractions, and the fate of dissolved nutrients. The 12% hillslope featured loamy sand A and E horizons overlying a sandy clay loam Bt at 1.25 m average depth. Interflow measured from two drains within an interception trench commenced after 131 and 208 mm of irrigation. Cumulative interflow equaled 49% of applied water. Conservative tracer differences between the collection drains indicated differences in flow paths and storages within the plot. Event water fractions rose steadily throughout irrigation, peaking at 50% sixteen h after irrigation ceased. Data implied that tightly held water exchanged with event water throughout the experiment and a substantial portion of preevent water was released from the argillic layer. Surface-applied dye tracers bypassed the matrix, with peak concentrations measured shortly after flow commencement, indicating preferential network conductivities of 864-2240 mm/h, yet no macropore flow was observed. Near steady-state flow conditions indicated average conductivities of 460 mm/h and 2.5 mm/h for topsoils and the Bt horizon, respectively. Low ammonium and phosphorus concentrations in the interflow suggested rapid uptake or sorption, while higher nitrate concentrations suggested more conservative transport. These results reveal how hydraulic conductivity variation and subsurface topographic complexity explain otherwise paradoxical solute and flow behaviors. Key Points:Interflow thresholds ranged from 131 to 208 mm, and flow paths were state-dependent Event-water fractions increased nearly linearly as irrigation progressed, reaching a maximum of 50% after 407 mm of irrigation Hillslope fluxes integrate spatiotemporal-varying processes controlled by conductivity, soil depth, and subsurface detention distributionsCorrespondence to: (2016), Interactions among hydraulic conductivity distributions, subsurface topography, and transport thresholds revealed by a multitracer hillslope irrigation experiment, Water Resour. Res., 52, 6186-6206,

A sprinkling experiment to quantify celerity–velocity differences at the hillslope scale

“…However, model calibrations on lateral subsurface flow and δ 2 H showed that higher bypass values produced better simulations (not shown); this is in contrast to McGuire et al (2007) who found that smaller bypass values produced better simulations. Our finding though is consistent with dye-staining experiments at a nearby site (McGuire et al, 2007), other studies at this site (McGuire and McDonnell, 2010;van Verseveld et al, 2008) and observations published in other studies (Radulovich et al, 1992;Hornberger et al, 1990;van Stiphout et al, 1987;Jackson et al, 2016) that all showed the importance of bypass flow. Additionally, our finding is in agreement with higher observed average vertical velocity at well E04 compared to the deep lysimeters during the sprinkler experiment.…”

“…They attributed this to the advancement of a pressure wave instead of advective flow of new water. Following recommencement of irrigation, Jackson et al (2016) observed during their multi-tracer hillslope irrigation experiment a fast trench flow and piezometer response indicating a pressure wave celerity much faster than observed dye tracer velocities. Rasmussen et al (2000) presented parametric expressions for the celerity, which predicted pressure wave travel times 2 to 15 times faster than the tracer velocity for their short-duration fluid irrigation experiments with intact saprolite columns.…”

“…To date, the focus of many experimental studies on tracer transport at the hillslope scale has been the qualitative and/or conceptual descriptions of velocities, observed transit times and dispersion of solutes (e.g., Anderson et al, 1997;Nyberg et al, 2003;Jackson et al, 2016). Studies interpreting the transport of solutes in a more quantitative way have largely relied on a lumped modeling approach (e.g., Rodhe et al, 1996;Jury and Sposito, 1985), assuming time-invariant flow conditions.…”

Interflow Is Not Binary: A Continuous Shallow Perched Layer Does Not Imply Continuous Connectivity

“…The 19.8‐m travel distance estimated for the Czech hillslope studied, for example, by Dusek and Vogel () closely matched the 25‐m contributing length estimated from their experimental work. Similarly, Jackson et al () also found that the contributing length estimated from the water balance for a plot‐scale irrigation experiment also closely matched the estimated DTD within the R watershed at the Savannah River Site. Irrigating a hillslope underlain by a fractured and leaky fragipan, Parlange et al () found that the effective contributing area was much smaller than the irrigated area.…”

“…We were also unable to use studies that reported large ranges of conductivities for each layer, but no central tendency. Hydraulic conductivities of surface soils and nearsurface impeding layers can range over several orders of magnitude (e.g., Jackson et al, 2016), and therefore, ranges of hydraulic conductivities without central tendencies could not be used to estimate DTD. 10.1029/2018WR022920 Detty and McGuire (2010).…”