Field estimates of groundwater circulation depths in two mountainous watersheds in the western U.S. and the effect of deep circulation on solute concentrations in streamflow

Abstract

Estimates of groundwater circulation depths based on field data are lacking. These data are critical to inform and refine hydrogeologic models of mountainous watersheds, and to quantify depth and time dependencies of weathering processes in watersheds. Here we test two competing hypotheses on the role of geology and geologic setting in deep groundwater circulation and the role of deep groundwater in the geochemical evolution of streams and springs. We test these hypotheses in two mountainous watersheds that have distinctly different geologic settings (one crystalline, metamorphic bedrock and the other volcanic bedrock). Estimated circulation depths for springs in both watersheds range from 0.6 to 1.6 km and may be as great as 2.5 km. These estimated groundwater circulation depths are much deeper than commonly modeled depths suggesting that we may be forcing groundwater flow paths too shallow in models. In addition, the spatial relationships of groundwater circulation depths are different between the two watersheds. Groundwater circulation depths in the crystalline bedrock watershed increase with decreasing elevation indicative of topography‐driven groundwater flow. This relationship is not present in the volcanic bedrock watershed suggesting that both the source of fracturing (tectonic versus volcanic) and increased primary porosity in the volcanic bedrock play a role in deep groundwater circulation. The results from the crystalline bedrock watershed also indicate that relatively deep groundwater circulation can occur at local scales in headwater drainages less than 9.0 km2 and at larger fractions than commonly perceived. Deep groundwater is a primary control on streamflow processes and solute concentrations in both watersheds.

The Importance of Capturing Topographic Features for Modeling Groundwater Flow and Transport in Mountainous Watersheds

“…For models with deep extinction depths ( d e ≥ 33.3 m), the spatial scaling of baseflow mean age and solute concentration is characterized by a positive slope (Figures h–l, o‐s, S5h–S5l, and S5o–S5s). This scaling pattern qualitatively matches the observed spatial scaling of solute concentration from watersheds with significant deep groundwater contribution to streamflow (e.g., Figures 7 and 8 in Frisbee et al, ) and the 3‐D catchment‐mixing streamflow generation conceptual model (see Figure 1 in Frisbee et al, ). For models with shallow extinction depths ( d e ≤ 10 m), the spatial scaling of baseflow mean age and solute concentration shows asymptotic behavior (Figures m and n, t and u, S5m and S5n, and S5t and S5u) and is consistent with observations in watersheds with dominant shallow flow contribution to streamflow (e.g., Figure 2 in Temnerud & Bishop, ) and 2‐D network‐mixing streamflow generation conceptual model (Figure 1 in Frisbee et al, ).…”

“…Although the model used by Vivoni et al () involves surface flow, vadose zone, and shallow groundwater flow processes, their modeling does not attempt to reproduce deep groundwater flow processes and the effect of capturing stream and ridge features on baseflow discharge at different scales. For some watersheds, deep groundwater flow has been found to be important for baseflow generation and stream chemistry evolution and is responsible for producing the simple emergent scaling relationship between groundwater contribution to stream flow and drainage area (discussed in detail in section 3.3; e.g., Frisbee et al, , ; Peralta‐Tapia et al, ). Thus, understanding how watershed topographic complexity affects groundwater contribution to streamflow across scales is important to building parsimonious watershed scale hydrological models (McDonnell et al, ) and improving hydrological predictions in data‐sparse regions (Sivapalan et al, ).…”

“…To test our hypothesis, we use the topography of the Rio Hondo watershed in Northern New Mexico, USA. This is a typical mountainous watershed with steep and complex topography characterized by significant contributions of deep groundwater to streamflow that tend to increase with drainage area (Frisbee et al, ; Tolley, ; Tolley et al, ). The hydrological and geochemical trends observed in this watershed are consistent with the conceptualization of a three‐dimensional, topography‐driven groundwater flow system (Tóth, , ) and a three‐dimensional catchment‐mixing watershed conceptual model (Frisbee et al, ).…”

“…Even though this simple conceptualization has limitations (Bresciani et al, ; Haitjema & Mitchell‐Bruker, ), it has proven useful to explain key processes in natural systems (Cardenas, ; Freeze & Witherspoon, ; Frisbee et al, , ; Genereux et al, ; Gomez & Wilson, ; Tóth, ; Wörman et al, ). For example, Tóth () used this assumption to explain the nesting of flow paths in a two‐dimensional, cross‐sectional basin model.…”

The Importance of Interflow to Groundwater Recharge in a Snowmelt‐Dominated Headwater Basin

“…• Supporting Information S1 Estimating recharge and groundwater discharge to streams is complicated by feedbacks between climate, vegetation, and topography, and by lack of data to characterize subsurface properties (Meixner et al, 2016), flow paths, and circulation depths (Frisbee et al, 2016). Estimating hydrologic partitioning is further hindered by the difficulty in quantifying snow distribution and snowmelt across the watershed (Deems et al, 2006;Harpold et al, 2012).…”