9.13 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles SM Wondzell, Olympia Forestry Sciences Laboratory, Olympia, WA, USA MN Gooseff, Pennsylvania State University, University Park, PA, USA

r 2013 Elsevier Inc. All rights reserved.

9.13.1 Introduction 204 9.13.2 The Effect of Geomorphology on HEFs 205 9.13.2.1 The Whole Network to Segment Scale 205 9.13.2.2 The Reach Scale – Setting the Potential for Hyporheic Exchange 206 9.13.2.2.1 Losing and gaining reaches 206 9.13.2.2.2 Changes in saturated cross-sectional area 208 9.13.2.3 The Subreach to Channel-Unit Scale – Hydrostatic Processes 209 9.13.2.3.1 Step–pool and pool–riffle sequences 209 9.13.2.3.2 Meander bends and point bars 210 9.13.2.3.3 Back channels and floodplain spring brooks 210 9.13.2.3.4 Secondary channels and islands 211 9.13.2.3.5 Spatial heterogeneity in saturated hydraulic conductivity 212 9.13.2.4 The Bedform Scale – Hydrodynamic Processes 212 9.13.2.5 The Particle Scale – Turbulent Diffusion 213 9.13.3 Discussion 213 9.13.3.1 Multiple Features Acting in Concert 213 9.13.3.2 Change in Processes Driving HEF through the Stream Network 214 9.13.4 Conclusion 215 References 215

Glossary area of floodplain alluvium through which hyporheic flow Bedform streambed features, such as ripples, sand-waves, occurs. or dunes, formed in streambed sediment through the Hyporheic exchange flow the flow of stream water from interaction between flowing water and sediment. Bedforms the surface stream channel, through the streambed and into can range in size from a few centimeters (ripples) to many the shallow unconfined aquifer beneath or adjacent to the meters (dunes). stream channel and back into the surface stream over Channel unit relatively homogeneous areas of a channel relatively short spatial (1-100s of meters) and temporal with characteristic substrate, depth, and flow pattern, (hours to weeks) scales. typically bounded by other channel areas with different Hyporheic zone thezoneofsaturatedsediments features. The length of channel units generally ranges underlying the surface stream or in the floodplain adjacent to from one to a few times the wetted width of the stream. the stream perfused with stream-source water that has recently Pools and riffles are common channel units in small left the stream channel and will return to the stream channel streams. relatively quickly. Stream water in hyporheic flow paths may Hydrodynamic exchange (also known as pumping mixwithgroundwatersothattherelativeproportionof exchange) exchange of stream water with the subsurface stream-source water in the hyporheic zone is highly variable, driven by the velocity head component of the total head ranging from 100% stream water to nearly 100% groundwater. gradient on the bed surface and exchange induced by The extent of the hyporheic zone can be arbitrarily defined on momentum gradients across beds and banks. the basis of the proportion of stream water present in the Hydrostatic exchange exchange of stream water with subsurface (e.g., 410%) or the residence time of stream water the subsurface driven by static hydraulic gradients (e.g., areas with residence times 24 h). that are determined by changes in water surface Stream reach a length of stream channel in which elevation, spatial heterogeneity in saturated hydraulic topographic features and sequences of channel units are conductivity, or changes in the saturated cross-sectional relatively homogeneous. The length of stream reaches are

Wondzell, S.M., Gooseff, M.N., 2013. Geomorphic controls on hyporheic exchange across scales: watersheds to particles. In: Shroder, J. (Editor in Chief), Wohl, E. (Ed.), Treatise on Geomorphology. Academic Press, San Diego, CA, vol. 9, Fluvial Geomorphology, pp. 203–218.

Treatise on Geomorphology, Volume 9 http://dx.doi.org/10.1016/B978-0-12-374739-6.00238-4 203 204 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

generally many 10s of times the width of the wetted stream stream stage (i.e., bank storage processes due to changes in channel. hydrostatic head gradients between stream and lateral Transient exchange the temporary movement of stream riparian aquifer). water into stream banks due to short-term increases in

Abstract

We examined the relationship between fluvial geomorphology and hyporheic exchange flows. We use geomorphology as a framework to understand hyporheic process and how these processes change with location within a stream network, and, over time, in response to changes in stream discharge and catchment wetness. We focus primarily on hydrostatic and hydrodynamic processes – the processes where linkages to fluvial geomorphology are most direct. Hydrostatic processes result from morphologic features that create elevational head gradients, whereas hydrodynamic processes result from the interaction between stream flow and channel morphologic features. We provide examples of the specific morphologic features that drive or enable hyporheic exchange and we examine how these processes interact in real stream networks to create complex subsurface flow nets through the hyporheic zone.

9.13.1 Introduction groundwater. Also, the residence time distribution of stream water in the hyporheic zone tends to be highly skewed, with Hyporheic exchange flow (HEF) is the movement of stream most of the stream water moving along short flow paths and water from the surface channel into the subsurface and back to thus having short residence times (hours), but some water the stream (Figure 1). Stream water in hyporheic flow paths either moving on long flow paths or encountering relatively may mix with groundwater so that the relative proportion of immobile regions having very extended residence times stream-source water in the hyporheic zone (HZ) is highly (weeks to months, or longer). The boundaries of the hypor- variable, ranging from 100% stream water to nearly 100% heic zone are arbitrary, commonly defined by the amount of stream-source water present in the subsurface. Triska et al. (1989) set a threshold of 10% stream-source water to define the limits of the hyporheic zone so that regions with o10% stream-source water were defined as groundwater. Alter- natively, the extent of the hyporheic zone can be delimited by water residence time, for example, the subsurface zone de- lineated by HEFs with residence times less than 24 h (the 24-h hyporheic zone; Gooseff, 2010). The objective of this chapter is to examine the relation Pool−riffle−pool sequence (a) between geomorphology and hyporheic processes. The two primary controls on hyporheic exchange are the gradients in Riffle total head established along and across streambeds and the hydraulic conductivity of the streambed and adjacent aquifer, Pool Pool both of which are significantly influenced by geomorphology. Total head (also known as potential) is the sum of pressure head, elevation head, and velocity head. Pressure head repre- sents height of a column of fluid to produce pressure. Velocity head represents the vertical distance needed for the fluid to fall freely (neglecting friction) to reach a particular velocity from rest. Elevation head represents the potential energy of a fluid Floodplain Riffle particle in terms of its height from reference datum. Hydro- Active Subsurface static head is referred to as the sum of elevation and pressure channel flow path head. Groundwater tables in unconfined aquifers represent Wetted the spatial gradients in hydrostatic head. A number of pro- (b) channel cesses either drive or enable HEF, several of which are based on changes in head gradients. We follow the organizational Figure 1 Idealized conceptual model of nested hyporheic flow paths structure presented by Ka¨ser et al. (2009), who divided these as influenced by step–pool or pool–riffle sequences. (a) Plan view processes into five distinct classes: showing arcuate HEF flow paths through the adjacent floodplain created by the change in the longitudinal gradient over the pool–riffle sequence where the amount of HEF is proportional to the head 1. Transient exchange – the temporary movement of stream gradient. (b) Longitudinal section along the thalweg of the stream water into stream banks due to short-term increases in showing the vertical component of HEF flows through the streambed. stream stage (i.e., bank storage processes due to changes in Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 205

hydrostatic head gradients between stream and lateral ripar- 9.13.2 The Effect of Geomorphology on HEFs ian aquifer; Lewandowski et al., 2009; Sawyer et al., 2009). 2. Turnover exchange – the trapping of stream water in the 9.13.2.1 The Whole Network to Segment Scale streambed during times of significant bed mobility (Elliott and Brooks, 1997b; Packman and Brooks, 2001). The geologic setting of the stream network is an important 3. Turbulent diffusion – exchange driven by slip velocity that is factor determining the likely occurrence of HEF, but there have created at the surface of the porous medium of the bed been few attempts to study HEF at this broad scale. Rather, our where streamwise velocity vectors continue to propagate into expectations are pieced together by drawing comparisons the surface layers of the bed (Packman and Bencala, 2000). among HZ studies that have been conducted in widely varying 4. Hydrostatic-driven exchange – exchange driven by static geologic settings, at different locations in the stream network, hydraulic gradients that are determined by changes in or under widely varying flow conditions. We expect that water surface elevation (Harvey and Bencala, 1993), spatial geomorphic–hyporheic relationships will differ substantially heterogeneity in saturated hydraulic conductivity, or among different geologic settings. changes in the saturated cross-sectional area of floodplain Fluvial geomorphic studies have examined the factors that alluvium through which hyporheic flow occurs. determine the types of channel morphologies present within 5. Hydrodynamic-driven exchange – exchange driven by the stream networks (Montgomery and Buffington, 1997; Wohl velocity head component of the total head gradient on the and Merritt, 2005; Brardinoni and Hassan, 2007). Mont- bed surface (i.e., pumping exchange; Elliott and Brooks, gomery and Buffington (1997) presented one such description 1997a, 1997b) and exchange induced by momentum gra- of the distribution of channel morphologies typical of many dients across beds and banks. mountainous landscapes. They showed that catchment area and channel longitudinal gradient controlled the develop- These classes of HEF processes are coupled to geomorphic ment of distinct channel types such that the channel types processes in many ways. This is most obvious for hydrostatic tended to follow a characteristic sequence within a catchment effects, which are directly dependent on channel and valley- (Figure 2(a)). In their example, this sequence starts with floor morphology and the depositional environment that bedrock and colluvial channels in the steepest, uppermost controls spatial heterogeneity in saturated hydraulic conduct- headwaters. As longitudinal gradients decrease, channels ivity (K). However, turnover of streambed sediment is also change to cascades, to step–pool, to plane–bed, to pool–riffle, related to fluvial geomorphic processes. Similarly, hydro- and the largest, lowest-gradient rivers were typified as dynamic effects result from the interaction of flow over stream dune–ripple channels. Along with these changes in channel bedforms. Geomorphic processes build stream bedforms and morphology, the following would be expected: decreased determine channel morphology, especially longitudinal gra- longitudinal gradient and mean grain size of streambed sedi- dient, bed roughness, and water depth, all of which influence ment, and increased depth, width, hydraulic radius, and flow velocity. The relationship between geomorphology and flow velocity (Leopold and Maddock, 1953; Wohl and the other classes of processes is less direct, but still plays a role Merritt, 2008). in controlling these processes through channel form and the In this chapter, we use Montgomery and Buffington’s size distribution of sediment that makes up the streambed. (1997) description of the sequence of channel types within This chapter focuses primarily on the hydrostatic and hydro- a catchment as a simple heuristic model to organize our dynamic processes where linkages to geomorphic processes examination of the relative importance of the different pro- are most direct. cesses that drive HEF within stream networks. We recognize We organize our discussion of the interactions between that local controlling factors commonly interrupt simple se- geomorphology and HEF using a hierarchical scaling frame- quencing of channel types. For example, landslides may block work developed for river networks (Frissell et al., 1986; large mainstem channels, creating locally steep gradients over Bisson and Montgomery, 1996), starting at the whole net- the landslide debris and uncharacteristically low gradients in work, through the stream segment, to the stream reach, to the depositional reach directly upstream (Benda et al., 2003). the channel unit, and down to the subchannel unit scale. We also recognize that regional differences in geology and We recognize that describing any given process or related geomorphology will lead to dramatically different spatial or- flow path at a single scale is somewhat arbitrary because of ganization of channel types (see, e.g., characteristic channel the nested structure of the hyporheic flow net and dispersion type in glaciated mountainous regions as described by Brar- among HEF flow paths. Despite this, the concept of scale is dinoni and Hassan (2007)). Our descriptions of the spatial an important heuristic tool to organize our understanding organization of stream types and the resulting HEF processes of hyporheic processes. In many senses, the reach scale is will have to be modified for any specific landscape. the most informative scale at which to consider HEF. A single Most hyporheic exchange results from head gradients reach, by definition, has characteristic channel morphology so pushing water through the streambed. The amount of stream that the factors driving HEF within the reach are relatively water entering the hyporheic zone is thus a function of the consistent. However, only a few of the geomorphic factors steepness of the head gradient and the saturated hydraulic driving HEF actually operate at this scale. Most of the drivers conductivity of the streambed and underlying aquifer. The work at the channel unit or smaller scales. Moreover, to head gradients can be induced in many ways, but the two of understand the importance of HEF in stream ecosystem pro- primary influence are the hydrostatic and hydrodynamic cesses, the cumulative effects of HEF must be evaluated at processes. The relative importance of each of these processes is scales much larger than a single reach. expected to vary among channel types and with longitudinal 206 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

Hill- slope Hollow

Colluvial- bedrock

Cascade Step− pool Plane− bed Pool−riffle Dune−ripple (a)

Hydrostatic contribution Hydrodynamic Turbulent contribution diffusion

Relative HEF Relative

(b) Longitudinal stream profile through stream network

Figure 2 (a) Hypothetical distribution of channel types along a stream profile in a mountainous stream catchment, and (b) the corresponding relative contribution of turbulent diffusion and both hydrostatic and hydrodynamic processes to the total amount of HEF occurring within a stream reach. (Note that boundaries between channel types are generally less distinct than shown here and that a range of conditions occurs within each category, thus the contribution of each process varies both among and within each channel type.) (a) Redrawn with permission from Montgomery, D.R., Buffington, J.M., 1997. Channel-reach morphology in mountain drainage basins. Geological Society of America Bulletin 109, 596–611. gradient. In high-gradient streams, channel forms, such as HEF occurs, and the potential effect of lateral groundwater step–pool sequences or pool–riffle sequences, can create inputs from adjacent hillslopes that might limit hyporheic very steep hydrostatic head gradients. Further, because of high expression. bed roughness and relatively shallow water depth, flow velocities tend to be lower in small steep streams than in 9.13.2.2.1 Losing and gaining reaches larger, low-gradient streams (Leopold and Maddock, 1953; Hyporheic exchange is likely to be more limited in strongly Wondzell et al., 2007). By contrast, it is difficult for natural gaining reaches than in neutral reaches because of steep processes to create steep changes in the longitudinal gradient streamward hydrologic gradients surrounding the channel in low-gradient streams. Instead, stream flow interacts with (Wroblicky et al., 1998; Storey et al., 2003; Malcolm et al., stream bedforms, such as dunes or ripples, such that hydro- 2003, 2005; Cardenas, 2009). Similarly, where water is lost dynamic forces dominate the development of head gradients to regional aquifers in strongly losing reaches, return flows through the streambed. Thus, we expect that hydrostatic effects of stream water back to the stream are likely to be severely will dominate in high-gradient channels and that hydro- restricted and thus also limit the expression of the hyporheic dynamic processes will dominate in low-gradient channels zone (Cardenas, 2009). These patterns of gains and/or losses (Figure 2(b)). Further, because channel types and longi- are controlled, at some level, by regional groundwater and tudinal gradients generally vary systematically within stream catchment characteristics interacting with smaller-scale effects. networks, we further expect that hydrostatic effects will tend In large gaining rivers, Larkin and Sharp (1992) demonstrated to dominate in the upper portions of stream networks and that the relative dominance of cross-valley versus down valley that the relative importance of hydrodynamic processes will flow paths through valley-floor aquifers varied depending on increase down the stream network. the longitudinal gradient of the valley floor and the hydraulic conductivity of the valley-floor alluvium. In higher-gradient reaches (40.004 m m–1) and in areas with coarser substrate, 9.13.2.2 The Reach Scale – Setting the Potential for flow was predominantly down valley. Conversely, where Hyporheic Exchange valley-floor gradients were shallower or sediment more finely The potential for HEF to occur varies within any given stream textured, flow tended to be toward the stream. Thus, the way reach. Roughly speaking, this potential is determined by in which lateral inputs influence hyporheic exchange is not the factors that generate head differences that drive HEF, solely a function of their magnitude, but also a function of the the properties of the subsurface alluvium through which ability of subsurface water to move down valley (Storey et al., Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 207

2003). The ratio between these two factors – the magnitude Simple generalizations of where and when lateral inputs of the inputs relative to down-valley flow – determines how will limit HEF are difficult because of the wide range of geo- hyporheic exchange is affected. morphic settings in which HEF occurs and because the mag- As a first approximation, the potential for down-valley nitude of lateral inputs changes with catchment wetness. flow can be estimated using the relationships summarized in Lateral inputs are expected to be high when catchments are Darcy’s law – that is, the product of the longitudinal valley wet and decrease as catchments dry out. However, lateral in- gradient, the saturated cross-sectional area of the floodplain puts are not spatially uniform. In steep mountainous settings, perpendicular to the direction of subsurface flow, and the the size of the upslope area draining directly to the valley floor hydraulic conductivity of the alluvium. As lateral inputs in- is important, concentrating lateral inputs in zones at the base crease, several factors may change: (1) water tables may rise, of hillslope hollows (Jencso et al., 2009). Lateral inputs may thus increasing the saturated thickness and the cross-sectional persist the entire year at the bases of the largest hillslope area through which water flows allowing the transmission of hollows. Most hillslope hollows are small, however, so that more water, or (2) flow paths may begin to turn obliquely most of the stream network would be disconnected from lat- toward the stream, which also increases the saturated cross- eral inputs except for short periods of time when catchments sectional area and may also increase head gradients. are very wet, for example, after large storms or during peak Under dry conditions when lateral inputs are relatively snowmelt. We are unaware of similar studies relating topo- small, the potential extent and magnitude of hyporheic ex- graphy to spatial patterns of hillslope inputs in areas of change can be fully expressed (Figure 3(a)). As subsurface low relief with humid climates. However, Storey et al. (2003) flows turn toward the channel, they begin to limit the extent reported that an extensive shallow surficial aquifer was present of the hyporheic zone with only a minor effect on the HEF along their lowland, low-gradient study reach and that lateral (Wondzell and Swanson, 1996). If sufficiently large, lateral inputs of groundwater substantially reduced both the extent inputs can severely limit both the spatial extent and magni- and the amount of HEFs except during summer base flow. tude of hyporheic exchange (Figure 3(b); Harvey and Bencala, Clearly, the influence of lateral inputs in lowland catchments 1993; Wroblicky et al., 1998; Storey et al., 2003; Cardenas and may be much different from that in steep mountainous Wilson, 2007; Malcolm et al., 2003; Soulsby et al., 2009). catchments. Changes in lateral inputs to streams do not occur in isol- ation. Rather, they are likely to be accompanied by corres- ponding changes in stream stage (and discharge). The change in water-table elevations resulting from changed lateral inputs must be considered relative to the accompanying changes in stream stage. Although the number of studies examining changes in hyporheic flow paths with changing catchment wetness is limited, studies in small mountain streams suggest that water-table elevations in the floodplain increase more than stream stage so that HEF is typically more restricted Higher-gradient, low lateral inputs when catchments are wet (Figures 4(a) and 4(b); Harvey and Bencala, 1993; Wondzell and Swanson, 1996; Stednick and (a) Fernald, 1999). Storey et al. (2003) reported similar results for a lowland, low-gradient river. In some cases, however, stream stage may change markedly without corresponding changes in precipitation recharge or changes in lateral inputs. Most examples of these processes come from large, lowland rivers because river stage is con- trolled by processes far upstream. These ‘bank storage’ pro- cesses (Pinder and Sauer, 1971) have been recognized as a form of transient hyporheic exchange (Figures 4(c) and 4(d)) Lower-gradient, high lateral inputs that can result from both in-bank or over-bank floods (Bates et al., 2000; Burt et al., 2002). In some situations, increased (b) stream stage may even lead to groundwater ridging in the Figure 3 Idealized conceptual model of the influence of lateral floodplain, reversing head gradients and limiting lateral inflows on hyporheic exchange flows. (a) A high-gradient stream groundwater inputs. Similarly, hyporheic exchange through where floodplain alluvium has relatively high saturated hydraulic stream banks can result from diel variations in stream stage conductivity under relatively dry conditions when lateral inputs are (and discharge) during snowmelt periods (Loheide and low and easily transported down valley via subsurface flow. Lateral Lundquist, 2009) or from tidally induced changes in water inputs still reach the stream, but are diverted toward zones with elevations in coastal streams and rivers (Bianchin et al., 2010). hyporheic upwelling. (b) A low-gradient stream where floodplain alluvium has relatively low saturated hydraulic conductivity under Transient hyporheic exchange may be especially evident in relatively wet conditions when lateral inputs are sufficiently large to regulated rivers where releases from dams (or other control overwhelm down-valley transport, causing lateral inputs to cross the structures) can result in large and rapid changes in river stage valley toward the stream. Lateral inputs severely restrict hyporheic without corresponding local precipitation to recharge flood- exchange flows. Legend follows Figure 1. plain aquifers (e.g., Fritz and Arntzen, 2007; Lewandowski 208 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

Precipitation

Stage increase Lateral inputs Lateral

(a) (b)

Stage increase Stage decrease

(c) (d) Figure 4 Idealized conceptual model of the influence changing stream stage on transient hyporheic exchange. (a, b) A losing reach at low base flow is converted to a gaining reach during a storm because precipitation recharge and lateral inputs of hillslope water increase water- table elevations more than the corresponding increase in the stream stage. The original stream and water table position from (a) is shown in (b) for reference (light-gray line). (c, d) An example of a river where changes in stream stage result from snowmelt, tidal influences, or dam releases far upstream. Increased stream stage causes stream water to flow into the adjacent aquifer creating a losing stream reach. Conversely, decreased stage leads to drainage of the aquifer creating a gaining reach. Alternating increases and decreases in stream stage lead to transient hyporheic exchange. The neutral condition (where stream stage is equal to the water-table elevation) is shown for reference (black-and-white dashed line). Legend follows Figure 1. et al., 2009; Sawyer et al., 2009; Francis et al., 2010). However, magnitude of HEF can be substantial, even in strongly gaining transient hyporheic exchange may not always result from reaches where the spatial extent of the hyporheic zone is fluctuations in river stage. For example, Hanrahan (2008) greatly restricted (Wondzell and Swanson, 1996; Cardenas studied vertical HEF through the streambed of a large, regu- and Wilson, 2007; Payn et al., 2009). lated gravel bed river where stage sometimes changed by nearly 2 m in an hour. For the most part, they did not observe transient hyporheic exchange related to changes in stage. They 9.13.2.2.2 Changes in saturated cross-sectional area concluded that hydrostatic and hydrodynamic processes re- The saturated cross-sectional area of the floodplain (orth- mained the dominant control on HEF. Notably, Hanrahan ogonal to groundwater flow path direction) is one of the (2008) did not examine lateral exchanges through the stream factors determining the amount of groundwater transmitted banks, which can be more responsive to changes in stage than down valley through the valley-floor alluvium. Thus, any are locations in the stream channel itself (Storey et al., 2003). change in the cross-sectional area along the length of a stream Water-table fluctuations in the floodplain at long distances reach will lead to parallel changes in the down-valley flow of from the stream are not necessarily indicative of extensive HEF water through the floodplain, thereby driving downwelling because pressure fluctuations can propagate through surficial from, or upwelling to, the stream (Stanford and Ward, 1993). (unconfined) aquifers much faster than does the actual flow of Downwelling occurs where valley floors increase in width, for stream water. This was clearly demonstrated by Lewandowski example, downstream of bedrock-constrained reaches et al. (2009) who showed that river water penetrated, at (Figure 5(b); Poole et al., 2004, 2006; Acuna and Tockner, most, only 4 m into the stream bank even though water- 2009). Conversely, upwelling occurs where valley floors table fluctuations were observed more than 300 m from narrow at the lower end of wide unconstrained reaches the river. (Figure 5(a); Baxter and Hauer, 2000; Acuna and Tockner, HEF can occur in strongly gaining and losing reaches be- 2009). Similarly, variations in the thickness of the surficial cause of the nested structure of hyporheic flow paths, and aquifer, caused by variations in depth to bedrock or other because HEF can occur at a variety of spatial scales. Thus, an confining layers, drive similar patterns of upwelling and envelope of the HZ can be set within larger nonhyporheic flow downwelling. For example, upwelling commonly occurs just paths (Figure 3(b); Cardenas and Wilson, 2007). Similarly, upstream of bedrock sills with a subsequent transition to smaller-scale HEF can occur as a result of smaller-scale geo- downwelling just downstream of such bedrock sills as the morphic drivers, even within a reach that is, overall, strongly surficial aquifer again thickens (Figure 5(b); Valett, 1993). losing (Payn et al., 2009). Further, because HEF is dominated This is easily observed in streams in arid regions during the dry by relatively near-stream flow paths that are short in length season, where perennial flow may only occur above bedrock and residence time (Kasahara and Wondzell, 2003), the sills, which force the subsurface flow to the surface. Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 209

longitudinal gradient and the saturated hydraulic conductivity control the amount of stream water exchanged with the subsurface. Many factors can modify the effect of steps or riffles on HEF. For example, the height of the step (or steepness of the Bedrock riffle) determines the head gradient available to drive HEF so Unconstrained stream reach gorge that a single very large step has the potential to drive more (a) HEF than if the same amount of elevational change is spread over several smaller steps (Kasahara, 2000). Because of this, large wood can be important in determining the amount of HEF in forest streams. Single logs tend to create frequent, small obstructions that collect and store small amounts of sediment, forming pool–step sequences in which the extent Bedrock sill (b) of the hyporheic zone tends to be small (Wondzell, 2006). Although logjams are less common, they can create large ob- Figure 5 Idealized conceptual model of the influence of the change structions storing sediment in wedges several meters deep and in saturated cross-sectional area of the floodplain on hyporheic Z exchange flows. (a) The influence of change in valley constraint with 10 m in length, and significantly widen constrained stream downwelling at the head of an unconstrained reach and upwelling at channels. Consequently, logjams can form extensive hypor- the downstream end of the reach caused by the transition from heic zones in steep, confined mountain streams (Wondzell, narrow bedrock gorges to wide alluvial valley floors. (b) The influence 2006). of variations in depth to bedrock forcing upwelling upstream of a Large, channel-spanning logs can wedge into steep bedrock sill and downwelling downstream, where the depth of narrow channels, forcing the accumulation of sediment in alluvium again increases. Legend follows Figure 1. channels, converting bedrock reaches to alluvial reaches with a step–pool morphology (Montgomery et al., 1996), thereby 9.13.2.3 The Subreach to Channel-Unit Scale – greatly enhancing HEF. Similarly, large wood can force plane– Hydrostatic Processes bed channels into a pool–riffle morphology (Montgomery et al., 1996), which should lead to more HEF than would Geomorphic features of the stream channel and valley floor be present in a comparable wood-free channel. Large wood within stream reaches control the elevation of surface water can have the opposite effect in channels that would have a and can thereby create significant head gradients through the free-formed pool–riffle morphology. In one documented case, valley-floor alluvium, driving HEF. Because these geomorphic accumulations of large wood tended to force a pool–riffle features are static on the timescales typical of hyporheic ex- channel toward a step–pool morphology (Wondzell et al., change (hours to weeks), they are broadly recognized as 2009). The channel adjusted to removal of all large, in-stream ‘hydrostatic processes’. wood by developing a better-defined pool–riffle structure around meander bends, leading to increased sediment storage. 9.13.2.3.1 Step–pool and pool–riffle sequences Continued channel adjustment over time, following the One of the best-studied examples of hydrostatic processes removal of large wood, eventually led to substantial increases involves the changes in water surface elevation along a in HEF. step–pool sequence and the resulting head gradients that drive The size, spacing, and sequence of channel units (e.g., HEF (Figure 1; Harvey and Bencala, 1993). Harvey and pools and riffles) along the stream longitudinal profile can Bencala (1993) showed that the change in the longitudinal also affect HEF (Anderson et al., 2005; Gooseff et al., 2006). gradient of the stream channel (which approximates the Anderson et al. (2005) made detailed measurements of stream energy profile) drove HEF. They also observed that HEF channel profiles and patterns of HEF, and showed that chan- flow paths tended to be curved – first curving away from nel unit size and spacing increased as did the length of the stream above the step or riffle and then curving back to channel characterized by downwelling with increasing drain- the stream below the step or riffle. Building from their ob- age area in a mountainous stream catchment. Gooseff servations, model analyses showed that along an idealized et al. (2006) built on these results, examining HEF using straight channel with homogeneous isotropic porous sedi- 2D groundwater models of idealized longitudinal profiles of ment, hyporheic flow paths around a change in the longi- mountain streams. The modeling results of Gooseff et al. tudinal gradient will exploit the full three-dimensional (3D) (2006) confirmed that both channel unit spacing and size saturated volume along the channel, thus extending both were important in determining hyporheic exchange patterns vertically beneath the streambed and horizontally through the of upwelling and downwelling. Perhaps more surprising, streambanks and near-stream aquifer (Figures 1(a) and 1(b)). however, was the observation that the sequence of channel Real streams are substantially more complicated, however, units also affected simulated HEF. Gooseff et al. (2006) such that changes in hydraulic conductivity of the alluvium, compared pairs of idealized stream reaches that varied only by bends in the channel, and the spatial location of lateral the way the longitudinal gradient changed over the pool–riffle groundwater inputs lead to the development of a complicated sequence – that is, the slope of the riffle was gradual on its flow net through the valley floor (e.g., Cardenas and Zlotnik, upstream end and steepest at its downstream end (described 2003). Despite these complexities, the steepness of the as a pool–riffle–step sequence) versus riffles that were initially hydraulic head gradient imposed by the change in the steep with the slope decreasing toward the downstream end 210 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

(described as a pool–step–riffle sequences). Simulated down- influenced by both the change in stream water elevation welling lengths were substantially longer for pool–riffle–step around the meander bend and the plan-view shape of the sequences than for pool–step–riffle sequences. meander bend. Highly evolved meander bends support steep head gradients across the mender neck because of the close proximity of the stream channels (Figure 6(b); Boano et al., 9.13.2.3.2 Meander bends and point bars 2006; Revelli et al., 2008) so that HEF is dominantly located in A variety of channel and valley-floor morphologic features, in the meander neck, with much reduced HEF across the re- addition to changes in the longitudinal gradient, create head mainder of the meander where head gradients are much lower. gradients with the potential to drive HEF. These include In other cases, meanders develop a characteristic pattern of channel meander bends and associated point bars, back alternating pools and riffles, with riffles located at the thalweg channels or floodplain spring brooks, and islands set between crossovers in the inflections between adjacent meanders and main and secondary channels. In all these cases, differences in pools or low-gradient runs wrapping around the point bar the elevational head of surface water between two channels, (Figure 6(c)). This combination of channel morphologic fea- between different points in a single channel around a meander tures can create complex HEF flow paths within meander bend, or between points on opposite sides of an island create bends. The residence times of HEF traversing meander bends head gradients that drive HEF. For example, head gradients can be quite short where meanders are small and saturated through the point bar in a meander bend are steeper than the hydraulic conductivities are high (Pinay et al., 2009). Con- longitudinal gradient of the stream channel around the point versely, residence times of HEF may be extremely long in me- bar (Peterson and Sickbert, 2006) so that stream water infil- ander bends of low-gradient rivers with fine-textured sediment trates the upper end of the point bar and is returned to the (Boano et al., 2006; Peterson and Sickbert, 2006). channel at the lower end of the point bar (Figure 6(a); Vervier and Naiman, 1992). More generally, these exchange flows occur across the full length of meander bends and are 9.13.2.3.3 Back channels and floodplain spring brooks Channel planforms are generally complex in wide floodplains, including a network of old or abandoned channels. If the upstream ends of these channels are plugged with sediment and if the downstream ends are sufficiently incised to intercept the water table and are connected back to the river at their downstream ends, they will act as drains, imposing head gradients from the stream to the old channel (Figure 7(a); Wondzell and Swanson, 1996; Poole et al., 2006). These channels are also known as floodplain spring brooks because (a) water upwells into the channel, forming a spring at its head. In addition to creating HEF, these channels will capture whatever water is in the surficial aquifer of the floodplain, including down-valley flows from upstream locations, and lateral inputs of groundwater or hillslope water from the valley margin. However, because lateral inputs tend to be small and spatially isolated (Jencso et al., 2009; and as discussed above), flood- plain spring brooks will most generally be fed by HEF (b) (Wondzell and Swanson, 1996; Jones et al., 2007). Abandoned channels can also be plugged at their down- stream ends and open to the river at their upstream ends. In this case, stream water can flow into the abandoned channel, infiltrate the channel bed, and raise the water table in the middle of the floodplain, thereby creating head gradients and driving HEF from the abandoned channel back to the main stream channel (Figure 7(b)). More complex situations arise Pool/Run when the longitudinal gradients in either the back channel or (c) mainstem channel are interrupted by steeper riffles or steps. Figure 7(c) shows the interactions between a back channel Figure 6 Idealized conceptual model of the influence of meander and riffle. Above the riffle, the elevation of water in the main bends on hyporheic exchange flow. (a) Simple, low-radius meander channel is higher than the back channel so water flows toward with HEF traversing the point bar and floodplain. (b) High-radius the spring brook. Downstream of the riffle, water in the main meander with incipient meander-cutoff, where the short distance channel is lower than the back channel so that the back across the neck leads to much higher head gradients and thus greater HEF through the neck than the remainder of the meander bar. channel loses water over its downstream extent, eventually (c) Meander bend with riffles located at the inflections between going dry before reaching the main channel. adjacent meanders so that head gradients through the point bar are The channel planform features that drive HEF can occur low and much of the HEF occurs around the riffles, driven by over a range of spatial scales, and their influence may change longitudinal changes in gradient. Legend follows Figure 1. through time as the stage height of water in the main channel Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 211

(a) (a)

(b) (b)

(c)C (c) Figure 8 Idealized conceptual model of the influence of mid-stream islands on hyporheic exchange flows. (a) Parallel and smooth Figure 7 Idealized conceptual model of the influence of back longitudinal gradients in the channels on both sides of the island channels on hyporheic exchange flows. (a) A back channel is incised create HEF flow paths that parallel stream flow. (b) Riffles at the below the water table, acts as a drain, and creates head gradients head of the island enhance head gradients leading to greater HEF. from the main channel to the back channel. (b) A back channel is (c) Offset riffles create strong cross-island head gradients and flow plugged near its downstream end, conducts water onto the paths, resulting in more HEF but with shorter flow path lengths and floodplain, raises the water table, and creates head gradients toward residence times. Legend follows Figure 1. the main channel. (c) Complex pattern of HEF caused by interactions between a riffle in the main channel and a back channel. its full length. The hyporheic hydrology of islands has not Paleochannels (dashed lines) support preferential flow. Legend been extensively studied. However, we expect that the surface follows Figure 1. water elevations in channels bounding the island create boundary conditions for total head and control HEF through changes. For example, a small gravel bar may have low points islands as is generally indicated by the available literature along the stream bank. At high stage, the entire gravel bar may (Dent et al., 2007; Francis et al., 2010). If channels along both be submerged. As the stage decreases, the center of the bar sides of the island are parallel and symmetric with constant may become exposed, creating a secondary channel along the longitudinal gradient, then flow through the island will par- bank. As the stage decreases further, flow may become dis- allel the channels and the head gradient driving flow will continuous through the secondary channel such that it func- equal the overall longitudinal gradient of the stream reach tions as a drain if it is plugged at the upstream end, or functions (Figure 8(a)). If riffles are present in the channels, the head as a conduit allowing stream water to infiltrate the surface of gradient through the island adjacent to the riffles can be much the gravel bar if it is plugged at its downstream end. Old steeper than the reach averaged longitudinal gradient channels in large floodplains may act similarly, with continu- (Figure 8(b)). Also, if riffles are displaced along the primary ous flow along their full length during floods, but becoming and secondary channels surrounding an elongated island such disconnected at intermediate to low stage, or even dry com- that a riffle is located near the head of the island in one pletely during periods of minimum discharge. In large flood- channel and near the tail of the island in the second channel, plain reaches, these channels can be hundreds of meters to the resulting head gradients would tend to drive flows laterally kilometers in length, extending nearly the full length of the through the island, leading to very large cross-sectional areas stream reach (Poole et al., 2006; Arrigoni et al., 2008). experiencing HEF, and therefore large amounts of HEF, albeit, with shorter-length flow paths (Figure 8(c)). Although islands 9.13.2.3.4 Secondary channels and islands may be uncommon in most channel types, they may domin- Islands present a special case of back channels in which the ate HEF in braided and anastomosing stream reaches (Ward channel is continuously connected to the main channel over et al., 1999; Arscott et al., 2001). Given the complexities 212 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles of potential sizes and shapes of islands and patterns in the paleochannel, or acting as distributaries routing water into longitudinal gradients in the bounding channels, the resulting the floodplain and imposing head gradients from the paleo- flow nets, residence times, and amounts of HEF are likely to channel to the stream. Locations of paleochannels are some- vary widely. times evident from shallow depressions along the floodplain. In other cases, over-bank deposition will have completely 9.13.2.3.5 Spatial heterogeneity in saturated hydraulic filled old channels so that there is no surficial indication on conductivity the flat floodplain surface. The influence of paleochannels is Fluvial processes control the depositional environment on the difficult to discern because networks of widely spaced wells are streambed and across the floodplain creating spatial hetero- unlikely to find and trace the location of these features along geneity in the texture of deposited and reworked sediment the length of the floodplain. As a consequence, their influence across a range of scales, from the surface of the streambed to on HEF has not been widely studied. the entire floodplain. Because sediment texture is closely re- lated to saturated hydraulic conductivity (K), these processes can substantially influence HEF. However, because of the 9.13.2.4 The Bedform Scale – Hydrodynamic Processes difficulties in quantifying these patterns at the scales at Channel hydraulics, and the spatial and temporal distribution which they influence HEF, they have been relatively little of velocity (kinetic energy) across streambeds, are significantly studied. At fine scales, streambed roughness can control the influenced by the form of the channel and the bedforms that depositional environment across the streambed (Buffington occur in channels. The continuous feedback between pressure and Montgomery, 1999), which lead to spatial patterns in the distribution and shear stress across the bed surface and the distribution of K within the streambed (Genereux et al., potential to erode the bed will cause turnover exchange to 2008), which in turn can influence both the location and occur during times of high flows. During lower flows, when amount of HEF. HEF will be restricted where the streambed is bed sediment is relatively stable, bedforms cause some level of clogged with fine sediment and preferentially located in zones form drag on the flows, inducing pressure distributions across with higher K. Experiments in flumes have also shown that the bedforms, thereby driving HEF at a scale smaller than the HEF can also influence patterns of fine-sediment deposition, bedform (Figure 9). The size of the bedform is set by both the with fine sediment preferentially deposited in downwelling energy regime of the reach and the material that makes up zones (Packman and MacKay, 2003; Rehg et al., 2005), which the reach, and the form drag induced on the water column by may explain differences in K between upwelling and down- the bedform is of course partly controlled by its size. Thus, the welling zones observed in a steep headwater stream (Scordo scale of HEF flowpaths induced by hydrodynamic exchange and Moore, 2009). across the bedforms will scale in part with the size of bedforms Spatially heterogeneous patterns in K influence HEF. For present (Cardenas et al., 2004). Finally, the heterogeneity example, groundwater flow modeling studies using homo- of the bed material that makes up the bedforms will have a geneous versus heterogeneous K showed that spatial hetero- distinct control on the flux rate and actual flowpaths through geneity may add substantial complexity to the spatial patterns and around the bedforms (Sawyer and Cardenas, 2009). of the hyporheic flow net (Woessner, 2000). Where relatively In sand-bed streams, hydrodynamic HEF has been exten- high K regions are aligned parallel with head gradients, they sively studied both theoretically and empirically. Typical create preferential flow pathways (Wagner and Bretschko, bedforms in sand-bed streams are dunes and ripples, which 2002) that can increase the total amount of HEF (Cardenas have a fairly predictable geometry and spacing, based on bed and Zlotnik, 2003; Cardenas et al., 2004). Results from sediment composition and flow rate. Thibodeaux and Boyle Cardenas et al. (2004) showed that influence of heterogeneity (1987) pioneered investigations of the hydrodynamic pressure in K was relatively greater in lower-gradient streams and where distribution across dunes, noting the penetration of channel head gradients driving HEF were reduced. To our knowledge, water into the porous bedforms. Further development of a the influence of fine-grained heterogeneity has not been stud- ‘pumping exchange’ model by Elliott and Brooks (1997a, ied in steeper channels where hydrostatic processes dominate. 1997b) expanded the ability to predict HEF and associated Fluvial processes also influence spatial patterns in K at the scale of the entire floodplain. Especially important is the layering of stream and floodplain alluvium. Layering can create strong vertical anisotropy (Chen, 2004), limiting Stream flow vertical exchange and promoting lateral flows through the streambed and floodplain (Packman et al., 2006; Marion et al., 2008). Overbank deposition can also bury back chan- nels creating ‘paleochannels’ where coarse streambed alluvium is buried under finer floodplain soils (Stanford and Ward, 1993; Stanford et al., 1994; Poole et al., 2004). If these paleochannels intercept the water table, they will function as Figure 9 Idealized longitudinal-section in the center of a straight large preferential-flow pathways that can route water the full stream channel with bedforms (triangular dunes) showing the length of a floodplain. In this regard, they function much like interaction with stream flow that creates regions of low and high a subsurface version of back channels or floodplain spring pressure on the streambed that drive HEF. Nonhyporheic subsurface brooks – either acting as drains lowering the water table in the flows, known as underflow (dashed arrows), are present beneath the floodplain and imposing head gradients from the stream to hyporheic zone. Legend follows Figure 1. Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 213 solute dynamics in channel–bed systems. Whereas most likely to be an important component of HEF in low-order, studies of hydrodynamic exchange processes were generally high-gradient streams (Figure 2(b)). Careful theoretical and carried out in or applied to flume studies, there has been at empirical research on turbulent diffusion has been conducted least one application of incorporating the pumping exchange largely on planar beds (Shimizu et al., 1990; Habel et al., model to tracer transport in field studies. Salehin et al. (2003) 2002). Therefore, in the complex bed topography of typical studied the transport of tracer along several kilometers of Sava gravel channels, turbulent diffusion will be a component of Brook in Sweden and successfully applied a solute transport HEF, likely not the singular driver of HEF. model to the observed data to explain long residence-time distributions using the pumping exchange model theory. The predictability of dune and ripple sizing and spacing makes the 9.13.3 Discussion pumping exchange model a useful tool to explore HEF in sand-bed streams and rivers. 9.13.3.1 Multiple Features Acting in Concert In gravel-bed streams, bedform types may be generally predictable (i.e., Montgomery and Buffington, 1997; Wohl In the examples presented above (Figures 1 and 3–9), we and Merritt, 2008; Chin, 2002), but the exact geometry and have mostly focused on single types of channel morphologic spacing of bedforms are less predictable, particularly at a scale features that drive or enable hydrostatic and hydrodynamic that will directly influence head distributions across and along HEF. However, these features never occur in isolation. Rather, a the channel. Hence, the velocity distribution in the channel single stream reach will typically contain many of the mor- and around the bedform, which contributes to hydrodynamic phologic features described above. Interactions among these exchange, is also unpredictable. Tonina and Buffington (2007) features are likely to be important in determining the actual conducted careful studies of total pressure distribution HEF in any given stream reach. In some cases, the effects across streambeds in flumes that had realistic geometry of a of multiple features could be additive and result in higher pool–riffle sequence in a gravel-bed channel. Their results HEF than if they did not co-occur. For example, cross-valley indicated that total head distribution (i.e., incorporating vel- flow paths between main channels and floodplain spring ocity head in addition to hydrostatic head) was important to brooks can be accentuated by riffles (Figure 7(c)). However, exchange at focused points in the channel where high velocity interactive effects could also cancel, for example, where riffles occurred. Further, they confirmed that, in general, there was at the inflection points of meander bends reduce head gradi- little or no contribution of velocity head to parts of the bed ents through point bars (Figure 6(c)). The interactions be- that were overlain by deeper, slower flow, and therefore a tween different processes driving or enabling HEF is complex, hydrostatic representation of exchange will likely be more and, to some degree, site specific, making it difficult to applicable in these locations. quantify the effects of these interactions. Because of these Regardless of the predictability of bedform geometry and difficulties, there are relatively few comparative studies that spacing, the associated hydrodynamic HEF may induce only have examined multiple processes concurrently, within natural limited lengths of exchange in the subsurface because much of stream channels, and attempted to evaluate the net effect of the exchange dynamics are expected to be vertical rather than each process on the total HEF within stream reaches. lateral. Exchange lateral to the channel is more likely to be Sensitivity analyses with groundwater flow models cali- driven by hydrostatic gradients set up across meander bends or brated to simulate HEF in a studied stream reach provide one bars (as described above). Hydrodynamic HEF will contribute opportunity to examine the relative importance of channel to, but be only one component of, total HEF in natural morphologic features on HEF where multiple features are channels, and its importance will be dictated by both channel present in a single reach. For example, Kasahara and Wondzell hydraulics and, if present, competing hydrostatic factors that (2003) examined a number of channel morphologic features can create steeper head gradients. among stream reaches of different sizes in a mountainous stream network under conditions of summer base flow dis- charge. In all cases, the single strongest driver determining the 9.13.2.5 The Particle Scale – Turbulent Diffusion amount of HEF occurring within the simulated stream reaches At the particle scale on streambeds, turbulent diffusion is was the change in longitudinal gradient over step–pool se- significantly influenced by the size and arrangement of surface quences in the second-order channel (Figure 10(a)) and sediment. Because turbulent diffusion is induced by the mo- pool–riffle sequences in the fifth-order channel. The shape of mentum transfer between the water column and the porous the hyporheic flow net in the fifth-order stream, however, was media, HEF due to turbulent diffusion is a function of the strongly controlled by the presence, location, and relative decreasing velocity profile within the surface layers of the elevation difference between water in the main channel and porous media (Shimizu et al., 1990). Thus, the distribution of the back channels (Figure 10(b)). Similarly, Cardenas et al. sediment at the surface will greatly influence the potential for (2004) examined sediment heterogeneity, size of bedforms, energy and mass transfer within this zone. Turbulent diffusion and both longitudinal and lateral head gradients in a low- HEF is prominent in gravel-bed streams where surface pores gradient, sand-bed stream. They found that HEF was greater are more likely to accommodate such open exchanges of where bedforms had higher amplitude and were more closely momentum across the bed (Tonina and Buffington, 2009). spaced. Spatial heterogeneity in K increased HEF relative to Beds composed of sand particle sizes and smaller provide too homogeneous simulations, as did inclusion of lateral head much resistance to the momentum exchange between the gradients, but the effect was small relative to the effect of the water column and the bed. Hence, turbulent diffusion is more size and spacing of bedforms. 214 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

S S S B I S S S S S S

Scale = 20 m (a) Log Bedrock M Wetted stream channel Valley-floor alluvium Back channel Hillslope or terrace Equipotential (0.2 m) Hyporheic flow path

M R B

T R R

I R R R

Scale = 50 m B (b) Figure 10 Examples of complex hyporheic flow paths resulting from interactions between channel morphologic features: (a) a steep, second- order step–pool channel with abundant large wood and (b) a moderate-gradient, fifth-order pool–riffle channel with two major spring brooks. Note the difference in spatial scale between the two stream reaches. Letters indicate morphologic features driving HEF: S, steps; R, riffles; M, meander bends; B, back channels/spring brooks; I, islands; and T, a steep riffle at the mouth of a tributary. Equipotential intervals (dashed lines) are 0.2 m. Hyporheic flow paths (arrows) are hand drawn to indicate general direction of hyporheic flow through the valley floor.

In low gradient channels, morphologic features can inter- within channels where the velocity head can provide add- act with changes in steam stage and lateral groundwater inputs itional potential and thereby influence the pattern of hypor- in ways that can substantially influence the amount of HEF heic exchange. over time, across seasons, or within a single storm event. Storey et al. (2003) examined HEF in a pool–riffle sequence 9.13.3.2 Change in Processes Driving HEF through the at both high- and low-base flow discharge. At high stage, the Stream Network stream tended to drown the riffle, substantially reducing the change in the longitudinal gradient over the pool–riffle Hyporheic exchange will vary widely across the sequence of sequence and thus reducing HEF. By contrast, at low stage, the channel types occuring in stream networks (Figure 2; Buf- water surface more closely followed the streambed topo- fington and Tonina, 2009). Channel networks generally fol- graphy, thus creating steeper head gradients that supported low a pattern of steep headwaters to low-gradient reaches more HEF. Storey et al. (2003) also showed that lateral inputs downstream. In mountain stream networks in particular, during the wet season were sufficient to eliminate most of the gradient changes are expected to be accompanied by channel HEF through the riffle. Cardenas and Wilson (2006) showed morphology changes resulting in a sequence of distinct that low rates of groundwater discharge limited the extent of channel morphologies (Figure 2(a)). Obviously, bedrock the HZ formed by the hydrodynamics of stream bedforms, reaches have negligible hyporheic zones (Gooseff et al., 2005; and that high rates of groundwater discharge could completely Wondzell, 2006). We are unaware of any studies of HEF in eliminate HEF. colluvial and cascade channel morphologies; however, the We know of only one study comparing the relative influ- extremely high longitudinal gradients of these channels likely ence of hydrostatic and hydrodynamic effects. In a flume, result in high-velocity underflow, which has been shown to Tonina and Buffington (2007) investigated the control of total restrict the extent of the hyporheic zone (Storey et al., 2003). head (i.e., including dynamic head) in driving hyporheic ex- In addition, the relatively disorganized structure of the bed change. Their results suggested that there are specific locations sediment prevents development of stepped water surface Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 215 profiles so that hydrostatically driven exchange due to longi- HEF in regulated rivers where stage fluctuates over daily cycles tudinal changes in gradient will likely be low. Turbulent dif- due to hydroelectric generation. Turbulent diffusion, on the fusion is likely to be a primary driver of HEF in such reaches other hand, is likely to occur in gravel-bed sections of the net- (Figure 2(b)). work, likely with the greatest potential influence in either cascade Free-formed step–pool channels occur at slightly lower or plane–bed sections of stream networks where stream velocities gradients (Figure 2(a)). These channels have well-organized are expected to be high. structure with periodic spacing of both steps and pools (Chin, 2002; Wohl and Merritt, 2008) that have been shown to be primary drivers of HEF (Figure 2(b); Kasahara and Wondzell, 9.13.4 Conclusion 2003). The addition of large wood can substantially increase sediment storage (Nakamura and Swanson, 1993; Mont- Hyporheic exchange results from distinct processes, and the gomery et al., 1996), the development of step–pool structure, relations between these processes and geomorphology are and the extent, amount, and residence times of HEF in these well understood from a mechanistic perspective. Thus, geo- stream reaches (Wondzell, 2006). Other hydrostatic factors morphology provides a critical framework to understand tend to have less dominance on HEF; these reaches have low hyporheic processes and how they change with location sinuosity so meander bends are uncommon and steep longi- within a stream network, and, over time, in response to tudinal gradients limit the potential for back channels to changes in stream discharge and catchment wetness. To the create lateral HEF flow paths. degree that these geomorphic patterns are predictable, they We are unaware of any published studies examining HEF in provide the foundation for hydrologists to make general pre- plane–bed channels. However, we expect HEF to be lower than dictions of the relative importance of the hyporheic zone at in either step–pool or pool–riffle channels (Figure 2(b)). The the scale of entire catchments. Reach-to-reach variability is streambed tends to be smoothly graded in these channels as high in stream networks, however, so understanding HEF at suggested by their name, and there is low spatial heterogeneity the reach scale continues to require detailed study of specific in surface texture (Buffington and Montgomery, 1999). Pools stream reaches. These studies are difficult and current meth- are widely spaced, and both steps and riffles are rare. Although odological approaches are insufficient to fully examine the full these channels occur as free-formed morphologies, pool–riffle suite of processes that account for patterns of HEF in any channels can be converted to plane–bed channels by land-use specific stream reach. Consequently, hyporheic studies tend to practices that increase sediment supply and through the direct focus on a single factor, or at most a small subset of the factors removal of large wood, with concurrent decreases in HEF. driving HEF. Hyporheic researchers recognize that such studies Lower in the stream network, channels tend have lower are incomplete. Detailed, holistic understanding of the im- longitudinal gradients (Figure 2(a)), and even in mountain- portance of different processes in driving HEF, how the relative ous areas, unconstrained stream reaches become increasingly importance of these processes changes with location in the common. Channel planforms can be quite complex in these stream network, with the specific structure of any given stream rivers and, as a consequence, a wide array of channel geo- reach, and with changes in discharge and lateral groundwater morphic features influences HEF. Braided and anastomosing inputs remains elusive. channels may form where sediment loads are high and stream banks are erodible; the complex of channels likely leads to substantial HEF through islands. Meandering channels form References under lower sediment loads and where banks are more stable. Meandering channels typically have pool–riffle morphologies, Acuna, V., Tockner, K., 2009. Surface–subsurface water exchange rates along although complexes of secondary channels, back channels, alluvial river reaches control the thermal patterns in an Alpine river network. and paleochannels are common, a legacy of past floods, Freshwater Biology 54, 306–320. channel avulsions, and overbank deposition. Because most Anderson, J.K., Wondzell, S.M., Gooseff, M.N., Haggerty, R., 2005. Patterns in stream HEF occurs along short, near-stream flow paths, riffles are the longitudinal profiles and implications for hyporheic exchange flow at the H.J. Andrews Experimental Forest, Oregon, USA. Hydrological Processes 19, 2931–2949. dominant feature determining the amount of HEF (Kasahara Arrigoni, A.S., Poole, G.C., Mertes, L.A.K., O’Daniel, S.J., Woessner, W.W., Thomas, and Wondzell, 2003). However, the shape of the hyporheic S.A., 2008. Buffered, lagged, or cooled? Disentangling hyporheic influences on flow net and the residence time distribution of HEF will be temperature cycles in stream channels. Water Resources Research 44, W09418. strongly influenced by the complex of channel planforms. http://dx.doi.org/10.1029/2007WR006480. Arscott, D.B., Tockner, K., Ward, J.V., 2001. Thermal heterogeneity along a braided Finally, hydrodynamic processes are expected to dominate floodplain river (Tagliamento River, Northeastern Italy). Canadian Journal of in streams with relatively mobile streambeds characterized by Fisheries and Aquatic Sciences 58, 2359–2373. dune–ripple bedforms. These streams have low longitudinal Bates, P.D., Stewart, M.D., Desitter, A., Anderson, M.G., Renaud, J.P., Smith, J.A., gradients and therefore channel morphologic features tend 2000. Numerical simulation of floodplain hydrology. Water Resources Research not to create steep hydrostatic head gradients (Figure 2(b)). 36, 2517–2529. Baxter, C.V., Hauer, F.R., 2000. Geomorphology, hyporheic exchange, and selection Other exchange processes are likely to be related to specific of spawning habitat by bull trout (Salvelinus confluentus). Canadian Journal of conditions. Turnover exchange will only occur when bed ma- Fisheries and Aquatic Sciences 57, 1470–1481. terial is mobile – a characteristic feature of both anastomosing Benda, L., Miller, D., Bigelow, P., Andras, K., 2003. Effects of post-wildfire erosion and dune–ripple channels. Transient exchange will only be ap- on channel environments, Boise River, Idaho. Forest Ecology and Management 178, 105–119. preciable during wet catchment conditions, when channel stage Bianchin, M., Smith, L., Beckie, R., 2010. Quantifying hyporheic exchange in a tidal is high and surrounding groundwater tables are comparatively river using temperature time series. Water Resources Research 46, W07507, low. However, transient exchange may be a dominant form of doi:10.1029/2009WR008365. 216 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

Bisson, P.A., Montgomery, D.R., 1996. Valley segments, stream reaches, and Harvey, J.W., Bencala, K.E., 1993. The effect of streambed topography on channel units. In: Hauer, R.R., Lamberti, G.A. (Eds.), Methods in Stream surface–subsurface water exchange in mountain catchments. Water Resources Ecology. Academic Press, New York, NY, ch. 2, pp. 23–52. Research 29, 89–98. Boano, F., Camporeale, C., Revelli, R., Ridolfi, L., 2006. Sinuosity-driven hyporheic Jencso, K.G., McGlynn, B.L., Gooseff, M.N., Wondzell, S.M., Bencala, K.E., exchange in meandering rivers. Geophysical Research Letters 33, L18406. http:// Marshall, L.A., 2009. Hydrologic connectivity between landscapes and streams: dx.doi.org/10.1029/2006GL027630. transferring reach- and plot-scale understanding to the catchment scale. Water Brardinoni, F., Hassan, M.A., 2007. Glacially-induced organization of channel-reach Resources Research 45, W04428. http://dx.doi.org/10.1029/2008WR007225. morphology in mountain streams. Journal of Geophysical Research 112, Jones, K.J., Poole, G.C., Woessner, W.W., et al., 2007. Geomorphology, hydrology, F03013. http://dx.doi.org/10.1029/2006JF000741. and aquatic vegetation drive seasonal hyporheic flow patterns across a gravel- Buffington, J.M., Montgomery, D.R., 1999. Effects of hydraulic roughness on dominated floodplain. Hydrological Processes 22, 2105–2113. surface textures of gravel-bed rivers. Water Resources Research 35, 3507–3521. Kasahara, T., 2000. Geomorphic controls on hyporheic exchange flow in mountain Buffington, J.M., Tonina, D., 2009. Hyporheic exchange in mountain rivers II: effects streams. M.S. thesis, Oregon State University, 103 pp. of channel morphology on mechanics, scales, and rates of exchange. Geography Kasahara, T., Wondzell, S.M., 2003. Geomorphic controls on hyporheic exchange Compass 3, 1038–1062. flow in mountain streams. Water Resources Research 39, 1005. http://dx.doi.org/ Burt, T.P., Bates, P.D., Stewart, M.D., Claxton, A.J., Anderson, M.G., Price, D.A., 10.1029/2002WR001386. 2002. Water table fluctuations within the floodplain of the River Severn, Ka¨ser, D.H., Binley, A., Heathwaite, A.L., Krause, S., 2009. Spatio-temporal England. Journal of Hydrology 262, 1–20. variations of hyporheic flow in a riffle–step–pool sequence. Hydrological Cardenas, M.B., 2009. Stream–aquifer interactions and hyporheic exchange in Processes 23, 2138–2149. gaining and losing sinuous streams. Water Resources Research 45, W06429. Larkin, R.G., Sharp, J.M., Jr., 1992. On the relationship between river-basin http://dx.doi.org/10.1029/2008WR007651. geomorphology, aquifer hydraulics, and ground-water flow direction in alluvial Cardenas, M.B., Wilson, J.L., 2006. The influence of ambient groundwater aquifers. Geological Society of America Bulletin 104, 1608–1620. discharge on exchange zones induced by current–bedform interactions. Journal Leopold, L.B., Maddock, T., Jr., 1953. The hydraulic geometry of stream channels of Hydrology 331, 103–109. and some physiographic implications. US Geological Survey Professional Paper Cardenas, M.B., Wilson, J.L., 2007. Exchange across a sediment–water interface No. 252, 57 pp. with ambient groundwater discharge. Journal of Hydrology 346, 69–80. Lewandowski, J., Lischeid, G., Nutzmann, G., 2009. Drivers of water level Cardenas, M.B., Wilson, J.L., Zlotnik, V.A., 2004. Impact of heterogeneity, bed fluctuations and hydrological exchange between groundwater and surface water forms, and stream curvature on subchannel hyporheic exchange. Water at the lowland River Spree (Germany): field study and statistical analyses. Resources Research, 40. http://dx.doi.org/10.1029/2004WR003008. Hydrological Processes 23, 2117–2128. Cardenas, M.B., Zlotnik, V.A., 2003. Three-dimensional model of modern channel Loheide, S.P., Lundquist, J.D., 2009. Snowmelt-induced diel fluxes through the bend deposits. Water Resources Research 39, 1141. http://dx.doi.org/10.1029/ hyporheic zone. Water Resources Research 45, W07404. 2002WR001383. Malcolm, I.A., Soulsby, C., Youngson, A.F., Hannah, D.M., 2005. Catchment-scale Chen, X.H., 2004. Streambed hydraulic conductivity for rivers in south-central controls on groundwater–surface water interactions in the hyporheic zone: Nebraska. Journal of the American Water Resources Association 40(3), implications for salmon embryo survival. River Research and Applications 21, 561–574. 977–989. Chin, A., 2002. The periodic nature of step–pool mountain streams. American Malcolm, I.A., Soulsby, C., Youngson, A.F., Petry, J., 2003. Heterogeneity in ground Journal of Science 302, 144–167. water–surface water interactions in the hyporheic zone of a salmonid spawning Dent, C.L., Grimm, N.B., Martı, E., Edmonds, J.W., Henry, J.C., Welter, J.R., 2007. stream. Hydrological Processes 17, 601–617. Variability in surface–subsurface hydrologic interactions and implications for Marion, A., Packman, A.I., Zaramella, M., Bottacin-Busolin, A., 2008. Hyporheic nutrient retention in an arid-land stream. Journal of Geophysical Research 112, flows in stratified beds. Water Resources Research 44, W09433. http://dx.doi.org/ G04004. http://dx.doi.org/10.1029/2007JG000467. 10.1029/2007WR006079. Elliott, A.H., Brooks, N.H., 1997a. Transfer of non-sorbing solutes to a streambed Montgomery, D.R., Abbe, T.B., Buffington, J.M., Peterson, N.P., Schmidt, K.M., with bed forms: theory. Water Resources Research 33, 123–136. Stock, J.D., 1996. Distribution of bedrock and alluvial channels in forested Elliott, A.H., Brooks, N.H., 1997b. Transfer of non-sorbing solutes to a streambed mountain drainage basins. Nature 381, 587–589. with bed forms: laboratory experiments. Water Resources Research 33, Montgomery, D.R., Buffington, J.M., 1997. Channel-reach morphology in mountain 137–151. drainage basins. Geological Society of America Bulletin 109, 596–611. Francis, B.A., Francis, L.K., Cardenas, M.B., 2010. Water table dynamics and Nakamura, F., Swanson, F.J., 1993. Effects of coarse woody debris on morphology groundwater–surface water interactions during filling and draining of a large and sediment storage of a mountain stream in western Oregon. Earth Surface fluvial island due to dam-induced river stage fluctuations. Water Resources Processes and Landforms 18, 43–61. Research 46, W07513, doi:10.1029/2009WR008694. Packman, A.I., Bencala, K.E., 2000. Modeling surface–subsurface hydrological Frissell, C.A., Liss, W.J., Warren, C.E., Hurley, M.D., 1986. A hierarchical framework interactions. In: Jones, J.B., Mulholland, P.J. (Eds.), Streams and Ground for stream classification: viewing streams in a watershed context. Environmental Waters. Academic Press, San Diego, CA. Management 10, 199–214. Packman, A.I., Brooks, N.H., 2001. Hyporheic exchange of solutes and colloids with Fritz, B.G., Arntzen, E.V., 2007. Effect of rapidly changing river stage on uranium moving bed forms. Water Resources Research 37, 2591–2606. flux through the hyporheic zone. Groundwater 45, 753–760. Packman, A.I., MacKay, J.S., 2003. Interplay of stream–subsurface exchange, clay Genereux, D.P., Leahy, S., Mitasova, H., Kennedy, C.D., Corbett, D.R., 2008. Spatial particle deposition, and streambed evolution. Water Resources Research 39, and temporal variability of streambed hydraulic conductivity in West Bear Creek, 1097. http://dx.doi.org/10.1029/2002WR001432. North Carolina, USA. Journal of Hydrology 358, 332–353. Packman, A.I., Marion, A., Zaramella, M., Chen, C., Gaillard, J.-F., Keane, D.T., Gooseff, M.N., 2010. Defining Hyporheic Zones Advancing Our Conceptual and 2006. Development of layered sediment structure and its effects on pore water Operational Definitions of Where Stream Water and Groundwater Meet. transport and hyporheic exchange. Water, Air, and Soil Pollution: Focus 6, Geography Compass 2, 1–11. 433–442. Gooseff, M.N., Anderson, J.K., Wondzell, S.M., LaNier, J., Haggerty, R., 2006. A Payn, R.A., Gooseff, M.N., McGlynn, B.L., Bencala, K.E., Wondzell, S.M., 2009. modeling study of hyporheic exchange pattern and sequence, size, and spacing Channel water balance and exchange with subsurface flow along a mountain of stream bedforms in mountain stream networks. Hydrological Processes 20, headwater stream in Montana, United States. Water Resources Research 45, 2443–2457. W11427. http://dx.doi.org/10.1029/2008WR007644. Gooseff, M.N., LaNier, J., Haggerty, R., Kokkeler, K., 2005. Determining in-channel Peterson, E.W., Sickbert, T.B., 2006. Stream water bypass through a meander neck, (dead zone) transient storage by comparing solute transport in a bedrock laterally extending the hyporheic zone. Hydrogeology Journal 14, 1443–1451. channel–alluvial channel sequence, Oregon. Water Resources Research 41, Pinay, G., O’keefe, T.C., Edwards, R.T., Naiman, R.J., 2009. Nitrate removal in the W06014. http://dx.doi.org/10.1029/2004WR003513. hyporheic zone of a salmon river in Alaska. River Research and Applications 25, Habel, F., Mendoza, C., Bagtzoglou, A.C., 2002. Solute transport in open channel 367–375. flows and porous streambeds. Advances in Water Resources 25, 455–469. Pinder, G.F., Sauer, S.P., 1971. Numerical simulation of flood wave modification due Hanrahan, T.P., 2008. Effects of river discharge on hyporheic exchange flows in to bank storage effects. Water Resources Research 7, 63–70. salmon spawning areas of a large gravel-bed river. Hydrological Processes 22, Poole, G.C., Stanford, J.A., Running, S.W., Frissell, C.A., 2006. Multiscale 127–141. geomorphic drivers of groundwater flow paths: subsurface hydrologic dynamics Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles 217

and hyporheic habitat diversity. Journal of the North American Benthological Tonina, D., Buffington, J.M., 2007. Hyporheic exchange in gravel bed rivers with Society 25, 288–303. pool–riffle morphology: laboratory experiments and three-dimensional modeling. Poole, G.C., Stanford, J.A., Running, S.W., Frissell, C.A., Woessner, W.W., Ellis, Water Resources Research 43, W01421. http://dx.doi.org/10.1029/ B.K., 2004. A Patch hierarchy approach to modeling surface and subsurface 2005WR004328. hydrology in complex flood-plain environments. Earth Surface Processes and Tonina, D., Buffington, J.M., 2009. Hyporheic exchange in mountain rivers I: Landforms 29, 1259–1274. mechanics and environmental effects. Geography Compass 3, 1063–1086. Rehg, K.J., Packman, A.I., Ren, J., 2005. Effects of suspended sediment Triska, F.J., Kennedy, V.C., Avanzio, R.J., Zellweger, G.W., Bencala, K.E., 1989. characteristics and bed sediment transport on streambed clogging. Hydrological Retention and transport of nutrients in a third-order stream in northwestern Processes 19, 413–427. California: hyporheic processes. Ecology 70, 1893–1905. Revelli, R., Boano, F., Camporeale, C., Ridolfi, L., 2008. Intra-meander hyporheic Valett, H.M., 1993. Surface–hyporheic interactions in a Sonoran Desert stream: flow in alluvial rivers. Water Resources Research 44, W12428. http://dx.doi.org/ hydrologic exchange and diel periodicity. Hydrobiologia 259, 133–144. 10.1029/2008WR007081. Vervier, P., Naiman, R.J., 1992. Spatial and temporal fluctuations of dissolved Salehin, M., Packman, A.I., Wo¨rman, A., 2003. Comparison of transient storage in organic carbon in subsurface flow of the Stillaguamish River (Washington, vegetated and unvegetated reaches of a small agricultural stream in Sweden: USA). Archiv fu¨r Hydrobiologie 123, 401–412. seasonal variation and anthropogenic manipulation. Advances in Water Wagner, F.H., Bretschko, G., 2002. Interstitial flow through preferential flow paths in Resources 26, 951–964. the hyporheic zone of the Oberer Seebach, Austria. Aquatic Sciences 64, Sawyer, A.H., Cardenas, M.B., 2009. Hyporheic flow and residence time distributions in heterogeneous cross-bedded sediment. Water Resources 307–316. Research 45, W08406. http://dx.doi.org/10.1029/2008WR007632. Ward, J.V., Malard, F., Tockner, K., Uehlinger, U., 1999. Influence of groundwater on Sawyer, A.H., Cardenas, M.B., Bomar, A., Mackey, M., 2009. Impact of dam surface water conditions in a glacial flood plain of the Swiss Alps. Hydrological operations on hyporheic exchange in the riparian zone of a regulated river. Processes 13, 277–293. Hydrological Processes 23, 2129–2137. Woessner, W.W., 2000. Stream and fluvial plain ground water interactions: rescaling Scordo, E.B., Moore, R.D., 2009. Transient storage processes in a steep headwater hydrogeological thought. Groundwater 38, 423–429. stream. Hydrological Processes 23, 2671–2685. Wohl, E., Merritt, D., 2005. Prediction of mountain stream geomorphology. Water Shimizu, Y., Tsujimoto, T., Nakagawa, H., 1990. Experiment and macroscopic Resources Research 41, W08419. http://dx.doi.org/10.1029/2004WR003779. modelling of flow in highly permeable porous medium under free-surface flow. Wohl, E., Merritt, D.M., 2008. Reach-scale channel geometry of mountain streams. Journal of Hydrosciences and Hydraulic Engineering 8, 69–78. Geomorphology 93, 168–185. Soulsby, C., Malcolm, I.A., Tetzlaff, D., Youngson, A.F., 2009. Seasonal and inter- Wondzell, S.M., 2006. Effect of morphology and discharge on hyporheic exchange annual variability in hyporheic water quality revealed by continuous monitoring flows in two small streams in the Cascade Mountains of Oregon, USA. in a salmon spawning stream. River Research and Applications 25, 1304–1319. Hydrological Processes 20, 267–287. Stanford, J.A., Ward, J.V., 1993. An ecosystem perspective of alluvial rivers: Wondzell, S.M., Gooseff, M.N., McGlynn, B.L., 2007. Flow velocity and the connectivity and the hyporheic corridor. Journal of the North American hydrologic behavior of streams during baseflow. Geophysical Research Letters Benthological Society 12, 48–60. 34, L24404. http://dx.doi.org/10.1029/2007GL031256. Stanford, J.A., Ward, J.V., Ellis, B.K., 1994. Ecology of the alluvial aquifers of the Wondzell, S.M., LaNier, J., Haggerty, R., Woodsmith, R.D., Edwards, R.T., 2009. Flathead River, Montana. In: Gibert, J., Danielopol, D.L., Stanford, J.A. (Eds.), Changes in hyporheic exchange flow following experimental wood removal in Groundwater Ecology. Academic Press, San Diego, CA, pp. 367–390. a small, low-gradient stream. Water Resources Research 45, W05406, 31. http:// Stednick, J.D., Fernald, A.G., 1999. Nitrogen dynamics in stream and soil waters. dx.doi.org/10.1029/2008WR007214. Journal of Range Management 52, 615–620. Wondzell, S.M., Swanson, F.J., 1996. Seasonal and storm dynamics of the Storey, R.G., Howard, K.W.F., Williams, D.D., 2003. Factors controlling riffle-scale hyporheic zone of a 4th-order mountain stream. I: hydrologic processes. Journal hyporheic exchange flows and their seasonal changes in a gaining stream: a of the North American Benthological Society 15, 1–19. three-dimensional groundwater flow model. Water Resources Research 39, 1034. Wroblicky, G.J., Campana, M.E., Valett, H.M., Dahm, C.N., 1998. Seasonal variation http://dx.doi.org/10.1029/2002WR001367. in surface–subsurface water exchange and lateral hyporheic area of two Thibodeaux, L.J., Boyle, J.O., 1987. Bed-form generated convective transport in stream–aquifer systems. Water Resources Research 34, 317–328. bottom sediment. Nature 325, 341–343.

Biographical Sketch

Steve Wondzell is a research riparian ecologist in the US Forest Service’s Pacific Northwest Research Station, located in Olympia, Washington, USA. His research broadly focuses on both basic and applied problems in watershed management and riparian and aquatic ecosystems. His basic research focuses on the interactions between hydrological, geomorphological, and ecological processes that create, maintain, or modify aquatic and riparian habitats, and the ways in which these processes either interact with, or are affected by, land-use practices. His applied research focuses on developing models and decision support tools that synthesize the current knowledge of aquatic and riparian systems into forms that can help inform management decisions at large spatial and temporal scales. 218 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles

Dr. Michael Gooseff is an associate professor in the Department of Civil & Environmental Engineering at Penn State University. He began his research career studying hyporheic zones associated with glacial meltwater streams in the McMurdo Dry Valleys of Antarctica for his PhD work at the University of Colorado. He continued to study hyporheic exchange at broader scales in his postdoctoral research in the HJ Andrews Experimental Forest of central Oregon, USA. He then moved to conducting hyporheic research in tundra streams of Alaska, mountain and valley streams in the intermountain western USA, and in restored and unrestored streams. He continues to have active research programs in both polar regions and continues strong collaborations to develop new tech- niques to study stream-groundwater interactions and implications.

TREATISE ON GEOMORPHOLOGY

EDITOR-IN-CHIEF John F. Shroder University of Nebraska at Omaha, Omaha, NE, USA

VOLUME 9 FLUVIAL GEOMORPHOLOGY

VOLUME EDITOR ELLEN WOHL Colorado State University, Fort Collins, CO, USA Academic Press is an imprint of Elsevier 32 Jamestown Road, London NW1 7BY, UK 225 Wyman Street, Waltham, MA 02451, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA

Copyright r 2013 Elsevier Inc. All rights reserved

The following chapters are US Government works in the public domain and not subject to copyright: 4.4 Nanoscale: Mineral Weathering Boundary 6.4 Karst Geomorphology: Sulfur Karst Processes 6.24 Cave Sediments as Geologic Tiltmeters 7.3 Stress, Deformation, Conservation, and Rheology: A Survey of Key Concepts in Continuum Mechanics 7.7 Surface Runoff Generation and Forms of Overland Flow 7.23 Mass-Movement Causes: Earthquakes 9.2 A River Runs Through It: Conceptual Models in Fluvial Geomorphology 9.9 Suspended Load 9.23 Waters Divided: A History of Alluvial Fan Research and a View of Its Future 9.29 Incised Channels: Disturbance, Evolution and the Roles of Excess Transport Capacity and Boundary Materials in Controlling Channel Response 9.36 Geomorphic Classification of Rivers 11.9 Loess and its Geomorphic, Stratigraphic, and Paleoclimatic Significance in the Quaternary 11.20 Anthropogenic Environments 12.2 Riverine Habitat Dynamics 12.7 Vegetation Ecogeomorphology, Dynamic Equilibrium, and Disturbance 12.8 The Reinforcement of Soil by Roots: Recent Advances and Directions for Future Research 12.21 Interactions among Hydrogeomorphology, Vegetation, and Nutrient Biogeochemistry in Floodplain Ecosystems

2.10 Hillslope Soil Erosion Modeling Copyright r 2013 R Brazier

2.12 Morphodynamic Modeling of Rivers and Floodplains Copyright r 2013 A Nicholas

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein, Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress

ISBN (print): 978-0-12-398353-4

For information on all Elsevier publications visit our website at books.elsevier.com

Printed and bound in Poland

13 14 15 16 17 18 10 9 8 7 6 5 4 3 2 1

Editorial: Claire Byrne Production: Karen East, Kirsty Halterman, Michael Nicholls CONTENTS

Preface xxvii Foreword xxix

VOLUME 9 Fluvial Geomorphology

9.1 Treatise on Fluvial Geomorphology E Wohl 1

Scales and Conceptual Models

9.2 A River Runs Through It: Conceptual Models in Fluvial Geomorphology GE Grant, JE O’Connor, and MG Wolman 6

Drainage Basin Processes and Analysis

9.3 Subsurface and Surface Flow Leading to Channel Initiation SK Kampf and BB Mirus 22 9.4 Network-Scale Energy Distribution P Molnar 43

Channel Processes

9.5 Reach-Scale Flow Resistance R Ferguson 50 9.6 Turbulence in River Flows T Buffin-Be´langer, AG Roy, and S Demers 69 9.7 The Initiation of Sediment Motion and Formation of Armor Layers EM Yager and HE Schott 87 9.8 Bedload Kinematics and Fluxes JK Haschenburger 103 9.9 Suspended Load RA Kuhnle 124 9.10 Bedforms in Sand-Bedded Rivers JG Venditti 137 9.11 Wood in Fluvial Systems AM Gurnell 163 9.12 Influence of Aquatic and Semi-Aquatic Organisms on Channel Forms and Processes JA Riggsbee, MW Doyle, JP Julian, R Manners, JD Muehlbauer, J Sholtes, and MJ Small 189

Exchanges/Fluxes

9.13 Geomorphic Controls on Hyporheic Exchange Across Scales: Watersheds to Particles SM Wondzell and MN Gooseff 203 9.14 Reciprocal Relations between Riparian Vegetation, Fluvial Landforms, and Channel Processes DM Merritt 219 9.15 Landslides in the Fluvial System O Korup 244

Channel Patterns

9.16 River Meandering JM Hooke 260 9.17 Morphology and Dynamics of Braided Rivers P Ashmore 289

v vi Contents

9.18 Hydraulic Geometry: Empirical Investigations and Theoretical Approaches BC Eaton 313 9.19 Anabranching and Anastomosing Rivers GC Nanson 330 9.20 Step–Pool Channel Features AE Zimmermann 346 9.21 Pool–Riffle DM Thompson 364

Fluvial Landforms

9.22 Fluvial Terraces FJ Pazzaglia 379 9.23 Waters Divided: A History of Alluvial Fan Research and a View of Its Future JD Stock 413

Paleohydrology

9.24 Quantitative Paleoflood Hydrology G Benito and JE O’Connor 459 9.25 Outburst Floods JE O’Connor, JJ Clague, JS Walder, V Manville, and RA Beebee 475 9.26 Global Late Quaternary Fluvial Paleohydrology: With Special Emphasis on Paleofloods and Megafloods VR Baker 511

Specific Fluvial Environments

9.27 Steep Headwater Channels M Church 528 9.28 Bedrock Rivers KX Whipple, RA DiBiase, and BT Crosby 550 9.29 Incised Channels: Disturbance, Evolution and the Roles of Excess Transport Capacity and Boundary Materials in Controlling Channel Response A Simon and M Rinaldi 574 9.30 Streams of the Montane Humid Tropics FN Scatena and A Gupta 595 9.31 Dryland Fluvial Environments: Assessing Distinctiveness and Diversity from a Global Perspective S Tooth 612 9.32 Large River Floodplains T Dunne and RE Aalto 645

Techniques of Study

9.33 Field and Laboratory Experiments in Fluvial Geomorphology E Wohl 679 9.34 Numerical Modeling in Fluvial Geomorphology TJ Coulthard and MJ Van De Wiel 694 9.35 Remote Data in Fluvial Geomorphology: Characteristics and Applications T Oguchi, T Wasklewicz, and YS Hayakawa 711

Management and Human Effects

9.36 Geomorphic Classification of Rivers JM Buffington and DR Montgomery 730 9.37 Impacts of Land-Use and Land-Cover Change on River Systems LA James and SA Lecce 768 9.38 Flow Regulation by Dams FJ Magilligan, KH Nislow, and CE Renshaw 794 9.39 Urbanization and River Channels A Chin, AP O’Dowd, and KJ Gregory 809 9.40 Impacts of Humans on River Fluxes and Morphology I Overeem, AJ Kettner, and JPM Syvitski 828 9.41 Geomorphologist’s Guide to Participating in River Rehabilitation GB Pasternack 843