The Role of Bedrock Groundwater in Rainfall∓Runoff Response At
Total Page:16
File Type:pdf, Size:1020Kb
Journal of Hydrology 450–451 (2012) 117–133 Contents lists available at SciVerse ScienceDirect Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol The role of bedrock groundwater in rainfall–runoff response at hillslope and catchment scales ⇑ C.P. Gabrielli a, , J.J. McDonnell b, W.T. Jarvis c,d a Department of Forest Engineering, Resources and Management, Oregon State University, Corvallis, OR 97331, USA b Global Institute for Water Security, National Hydrology Research Centre, University of Saskatchewan, 11 Innovation Boulevard, Saskatoon, SK, Canada S7N 3H5 c Institute for Water and Watersheds, Oregon State University, Corvallis, OR 97331, USA d Department of Geosciences, Oregon State University, Corvallis, OR 97331, USA article info summary Article history: Bedrock groundwater dynamics in headwater catchments are poorly understood and poorly character- Received 17 November 2011 ized. Direct hydrometric measurements have been limited due to the logistical challenges associated Received in revised form 2 May 2012 with drilling through hard rock in steep, remote and often roadless terrain. We used a new portable bed- Accepted 9 May 2012 rock drilling system to explore bedrock groundwater dynamics aimed at quantifying bedrock groundwa- Available online 18 May 2012 ter contributions to hillslope flow and catchment runoff. We present results from the Maimai M8 This manuscript was handled by Philippe Baveye, Editor-in-Chief, with the assistance research catchment in New Zealand and Watershed 10 (WS10) at the H.J. Andrews Experimental Forest of M. Todd Walter, Associate Editor in Oregon, USA. Analysis of bedrock groundwater at Maimai, through a range of flow conditions, revealed that the bedrock water table remained below the soil–bedrock interface, indicating that the bedrock Keywords: aquifer has minimal direct contributions to event-based hillslope runoff. However, the bedrock water Hillslope hydrology table did respond significantly to storm events indicating that there is a direct connection between hill- Rainfall runoff slope processes and the underlying bedrock aquifer. WS10 groundwater dynamics were dominated by Bedrock groundwater fracture flow. A highly fractured and transmissive zone within the upper one meter of bedrock conducted Stormflow rapid lateral subsurface stormflow and lateral discharge. The interaction of subsurface stormflow with bedrock storage directly influenced the measured hillslope response, solute transport and computed mean residence time. This research reveals bedrock groundwater to be an extremely dynamic component of the hillslope hydrological system and our comparative analysis illustrates the potential range of hydro- logical and geological controls on runoff generation in headwater catchments. Ó 2012 Elsevier B.V. All rights reserved. 1. Introduction nia showed the potentially significant hydrologic influence of underlying bedrock in rainfall–runoff delivery at the hillslope Process understanding of the rainfall–runoff response of steep scale. They found that stormflow followed fracture pathways with- hillslopes and headwater catchments has evolved greatly since in the shallow weathered bedrock and interacted with the overly- the foundational work of Hursh (1936). Early reviews (Dunne, ing colluvium when flows were forced upwards by more 1978) and more recent reviews (Bonell, 1998; Bachmair and Weil- competent bedrock, creating zones of transient saturation. Later er, 2011) have chronicled the development of concepts on rapid work at the Coos Bay Oregon site by Montgomery et al. (1997) subsurface stormflow development and the integration of individ- and Anderson et al. (1997) also noted subsurface flow paths that ual hillslope responses that create the integrated catchment re- traversed the soil and bedrock zones in steep unchanneled slopes. sponse. Despite extensive research revealing dominant processes Water exfiltrating from the bedrock during storm events and sprin- in different environments, the majority of the work to date has fo- kling experiments produced perched transient water tables at the cused exclusively on lateral flow in the soil mantle (Buttle, 1998; soil–bedrock interface that influenced directly, subsurface storm- Hewlett and Hibbert, 1967; McDonnell, 1990; Tsuboyama et al., flow and slope instability. Additionally, their bromide tracer injec- 1994). tions showed rapid movement of bedrock flow to the catchment Despite a largely soil-centric view of hillslope hydrology, early outlet identifying the importance of bedrock flow paths for hill- work by Wilson and Dietrich (1987) in a zero order basin in Califor- slope-to-catchment integration. More recently, Kosugi et al. (2008), building upon other important work in Japan (Katsuyama et al., 2005; Onda et al., 2001; Uchida et al., 2003), showed the importance of bedrock groundwater in a granitic catchment in ⇑ Corresponding author. Tel.: +1 603 568 6187. Central Japan. They found that transient saturation at the soil bed- E-mail addresses: [email protected] (C.P. Gabrielli), jeff.mcdonnell@ orst.edu (J.J. McDonnell), [email protected] (W.T. Jarvis). rock interface was connected to the rise and fall of deeper bedrock 0022-1694/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhydrol.2012.05.023 118 C.P. Gabrielli et al. / Journal of Hydrology 450–451 (2012) 117–133 groundwater that ultimately influenced the chemical, spatial and relation to the role of bedrock groundwater in rainfall–runoff re- temporal characteristics of subsurface water movement into the sponse at hillslope and catchment scales: stream channel. In some bedrock aquifers, fracture flow is the key feature con- A. How do hydrogeological features and the structure of bed- trolling bedrock groundwater contributions to hillslope flow and rock influence bedrock groundwater movement at the hill- catchment runoff response. Fracture flow through bedrock is con- slope scale? trolled by fracture network density, geometry and connectivity B. How does bedrock groundwater react to storm events? (Banks et al., 2009) and can be extremely complex and heteroge- C. Does bedrock groundwater contribute directly to hillslope neous. Fracture zones separated by competent bedrock may create discharge through exfiltration at the soil–bedrock interface? compartmentalized aquifers, while faulting, weathering and other large scale geologic processes may help induce connectivity be- We address these questions by combining standard hydromet- tween fracture pathways (Dietrich et al., 2005). Haria and Shand ric approaches that include wells, hillslope trenches and gauging (2004) found complex flow processes at depth-specific horizons stations, with the new Gabrielli and McDonnell (2011) drilling sys- in fractured bedrock in the riparian and lower hillslope region of tem and stable isotope tracer analysis of storm event streamflow, the Hafren catchment in Plynlimon, Wales. They observed that hillslope trench water and bedrock groundwater. We targeted nine the dual-porosity environment of fractured bedrock promoted ra- and 16 week field campaigns during the wet season at the Maimai pid storm response and minimal storage on a storm event time and H.J. Andrews, respectively. scale. Despite growing awareness of the potential significance of bed- rock groundwater at hillslope and catchment scales, there still re- 2. Study sites mains a very limited number of studies that have monitored hillslope groundwater in competent and fractured bedrock Two well-established benchmark experimental hillslopes and (McDonnell and Tanaka, 2001). Access remains the key logistical catchments were investigated: Watershed M8 (M8) at the Maimai hurdle limiting studies in characteristically steep, unstable, and of- in New Zealand and watershed 10 (WS10) at the H.J. Andrews in ten roadless headwater terrain. Where access has been limited, Oregon, USA. These watersheds appear nearly identical in many re- many studies have inferred catchment groundwater dynamics spects including size, physical hillslope characteristics and rain- through intensive studies of spring discharge, rather than direct fall–runoff characteristics (Table 1). They differ significantly, measurements taken within the bedrock itself (Iwagami et al., however, in their underlying bedrock geology. WS10 is dominated 2010; Katsuyama et al., 2005; Uchida et al., 2003). While useful, by layers of fractured pyroclastic tuff and breccias (Swanson and the black box nature of spring studies limit our ability to mechanis- James, 1975). M8 is underlain by a firmly-compacted conglomerate tically assess the dynamics of internal bedrock groundwater and its with little or no fracturing (O’Loughlin et al., 2002). Additionally, connection to hillslope processes in the soil mantle. Indeed, in hill- slope and catchment hydrology we struggle to know the general Table 1 involvement, if at all, of bedrock groundwater in forming satura- M8 and WS10 hillslope and catchment physical, meteorological and process tion at the soil–bedrock interface where large anisotropy and rapid characteristics. Source: Sayama and McDonnell (2009) as summarized from primary generation of lateral subsurface stormflow has been widely ob- literature from the two sites. served (Weiler et al., 2006). We lack, particularly at previously Characteristic M8, Maimai WS10, H.J. Andrews well-monitored and well-documented sites, an understanding of Size (ha) 3.8a 10.2 how bedrock groundwater couples to rapid event runoff genera- Slope angle (°)34 29 tion and flow