Bulk density optimization to determine subsurface hydraulic properties in Rocky Mountain catchments using the GEOtop model

Abstract

Integrated watershed models can be used to calculate streamflow generation in snow‐dominated mountainous catchments. Parameterization of water flow is often complicated by the lack of information on subsurface hydraulic properties. In this study, bulk density optimization was used to determine hydraulic parameters for the upper and lower regolith in the GEOtop model. The methodology was tested in two small catchments in the Dry Creek Watershed in Idaho and the Libby Creek Watershed in Wyoming. Modelling efficiencies for profile‐average soil–water content for the two catchments were between 0.52 and 0.64. Modelling efficiencies for stream discharge (cumulative stream discharge) were 0.45 (0.91) and 0.54 (0.94) for the Idaho and Wyoming catchments, respectively. The calculated hydraulic properties suggest that lateral flow across the upper–lower regolith interface is an important driver of streamflow in both the Idaho and Wyoming watersheds. The overall calibration procedure is computationally efficient because only two bulk density values are optimized. The two‐parameter calibration procedure was complicated by uncertainty in hydraulic conductivity anisotropy. Different upper regolith hydraulic conductivity anisotropy factors had to be tested in order to describe streamflow in both catchments.

Quantifying contributions of snowmelt water to streamflow using graphical and chemical hydrograph separation

“…We make no distinction between surface (i.e., overland flow) and subsurface streamflow origins in the nomenclature of our selected hydrograph components based on extensive field observations. During peak snowmelt conditions, water stored in hillslopes was rapidly transported via lateral flowpaths arising from macropores and/or interfaces between depositional units with large density contrasts (Fullhart et al, 2019; McNamara et al, 2005; Roberge & Plamondon, 1987; Uchida et al, 2005) indicative of threshold response (Penna et al, 2011; Spence, 2007; Tromp‐Van Meerveld & McDonnell, 2006). This quick moving lateral flow was observed in NONM100 with low SC, exfiltrating from the base of hillslopes where fully saturated riparian areas could not accommodate additional subsurface flow (Figure S4).…”

“…A porosity transition within the regolith indicated by seismic refraction surveys suggests the top layer of upper regolith has relatively high hydraulic conductivity capable of draining quickly. A novel bulk density optimisation method performed in NONM100 supports the decline in porosity with depth, which creates conditions that favour shallow lateral flow through the upper regolith as a dominant streamflow generation mechanism (Fullhart et al, 2019).…”

Hydrogeophysical Inversion of Time‐Lapse ERT Data to Determine Hillslope Subsurface Hydraulic Properties

“…Recently, Engda et al. (2016) employed a fine‐earth fraction optimization method that was successful in determining hydraulic properties in mountainous environments and has shown promise for reducing the overall number of parameters needed for calibration (Fullhart et al., 2018, 2019). The fine‐earth fraction optimization method relates measured fine‐earth characteristics (texture and gravimetric water retention) to a volumetric water retention function through a fine‐earth fraction bulk density whose value is calibrated to implicitly account for varying fractions of fine‐earth and coarse fragments in the subsurface.…”

Quantifying daily versus annual contributions of snowmelt water to streamflow using graphical and geochemical hydrograph separation

“…A porosity transition within the regolith indicated by seismic refraction surveys suggests the top layer of upper regolith has relatively high hydraulic conductivity capable of draining quickly. A novel bulk density optimization method performed in NONM100 supports the decline in porosity with depth, which creates conditions that favor shallow lateral flow through the upper regolith as a dominant streamflow generation mechanism (Fullhart, Kelleners, Chandler, McNamara, & Seyfried, 2019).…”

“…overland flow) and subsurface runoff in the nomenclature of our selected hydrograph components based on extensive field observations. During peak snowmelt conditions, water stored in hillslopes was rapidly transported via lateral flowpaths arising from macropores and / or interfaces between depositional units with large density contrasts (Fullhart et al, 2019;McNamara, Chandler, Seyfried, & Achet, 2005;Roberge & Plamondon, 1987;Uchida, Tromp-Van Meerveld, & McDonnell, 2005;) indicative of threshold runoff response (Penna, Tromp-Van Meerveld, Gobbi, Borga, & Dalla Fontana, 2011;Spence, 2007;Tromp-Van Meerveld & McDonnell, 2006). This quick moving lateral runoff was observed in NONM100 with low SC, exfiltrating from the base of hillslopes where fully saturated riparian areas could not accommodate additional subsurface flow (Figure S4).…”