Elsevier

Geomorphology

Volume 80, Issues 3–4, 30 October 2006, Pages 295-309
Geomorphology

Influence of vertical channel change associated with wood accumulations on delineating channel migration zones, Washington, USA

https://doi.org/10.1016/j.geomorph.2006.03.002Get rights and content

Abstract

We combine hydraulic modeling and field investigations of logjams to evaluate linkages between wood-mediated fluctuations in channel-bed-and water-surface elevations and the potential for lateral channel migration in forest rivers of Washington state. In the eleven unconfined rivers we investigated, logjams were associated with reduced channel gradient and bank height. Detailed river gauging and hydraulic modeling document significant increases in the water-surface elevation upstream of channel-spanning wood accumulations. Logjams initiated lateral channel migration by increasing bed-or water-surface elevations above adjacent banks. Because the potential for a channel to avulse and migrate across its floodplain increases with the size and volume of instream wood, the area of the valley bottom potentially occupied by a channel over a specified timeframe — the channel migration zone (CMZ) — is dependent on the state of riparian forests. The return of riparian forests afforded by current land management practices will increase the volume and caliber of wood entering Washington rivers to a degree unprecedented since widespread clearing of wood from forests and rivers nearly 150 years ago. A greater supply of wood from maturing riparian forests will increase the frequency and spatial extent of channel migration relative to observations from wood-poor channels in the period of post-European settlement. We propose conceptual guidelines for the delineation of the CMZs that include allowances for vertical fluctuations in channel elevation caused by accumulations of large woody debris.

Introduction

The interaction of large woody debris (LWD) with water and sediment can profoundly affect channel processes and form. Instream wood can change local sediment transport capacity and supply by increasing hydraulic roughness and impounding sediment behind logjams (Keller and Swanson, 1979, Harvey et al., 1987, Shields and Gippel, 1995, Gippel et al., 1996, Montgomery et al., 1996, Buffington and Montgomery, 1999, Manga and Kirchner, 2000). Sediment storage and pool formation associated with logjams can force alluvial morphologies in otherwise bedrock reaches of mountain channels (Montgomery et al., 1996, Montgomery et al., 2003). The sediment capacitance provided by many small woody debris dams in confined channels can moderate sediment flux from small basins impacted by punctuated sediment inputs (Massong and Montgomery, 2000, Lancaster et al., 2001). Larger, unconfined river systems are capable of organizing LWD into stable structures that can armor banks against erosion or redirect flow into banks (Harmon et al., 1986, Nakamura and Swanson, 1993, Abbe and Montgomery, 1996, Abbe and Montgomery, 2003). Local aggradation behind stable logjams can raise the bed-surface elevation of channels and increase the potential for floodplain inundation and lateral channel migration (Abbe and Montgomery, 2003). Consequently, changes in the age structure and composition of riparian forests can dramatically influence channel dynamics and the potential for lateral channel migration of unconfined rivers.

Archival and field investigations of floodplain forests in the Puget Lowland of Washington state indicate that wood had a dominant influence on riverine processes and aquatic habitat until late in the mid-nineteenth century, prior to extensive European settlement (Collins et al., 2002). Written accounts by settlers document recurrent flooding and channel avulsion attributed to the chronic clogging of rivers by abundant logjams. In one of the earliest published accounts, Wolff (1916, p. 453) described the formation of logjams as a principal control on the position of the lower White River near Auburn, Washington:

… the river has deviated greatly from the shortest course, the normal bed and in places has covered considerable areas, cutting away the land and leaving gravel bars below flood level. A close study of conditions shows that in every instance the current was first deflected by an accumulation of drift, the huge timber of this section serving readily in its formation. When one of these catches on an obstruction below, it quickly entangles others, and the mass of drift thus formed is solid enough to deflect the current. Gravel, sand, and silt collect in the dead water, behind the drift piles, strengthening them and preventing the river from returning to its original bed. Evidences of this action are plentiful, and, in the narrow valley of the upper reaches, show that the river has been forced from the hills on one side to those of the other, a distance of ½ mile or more, and the original bed has become overgrown with very heavy timber.

Such inconveniences to navigation and floodplain settlement led to the common practice of clearing snags from Puget Sound rivers throughout the late-nineteenth and early twentieth centuries (Collins et al., 2002). Riparian deforestation and levee construction reduced wood recruitment rates and wood abundance in western Washington rivers (Beechie et al., 2001). But in unconfined rivers retaining mature floodplain forest, large logjams still form and can significantly impact flooding and channel migration (Collins and Montgomery, 2001, Collins and Montgomery, 2002, Abbe et al., 2003).

Accumulations of LWD in rivers are likely to become increasingly common with the adoption of regulatory guidelines to protect riparian forests, habitat restoration efforts that re-introduce wood to channels and re-connect floodplains with levee setbacks, and the return of normative flows in the wake of dam-removal projects. The volume of wood entering Washington rivers will be unprecedented since historical forest clearing and instream wood removal by early settlers nearly 150 years ago. In Washington state, guidelines for the delineation of channel migration zones (CMZs) — the area potentially affected by the movement of a channel across its valley bottom over a specified timescale — rely heavily on the historical record of previous channel migration (e.g., Washington Forest Practices Board (WFPB), 2001, Rapp and Abbe, 2003). However, the aerial photographic record in most areas does not begin until the 1930s, well after instream wood removal and widespread harvest of low-elevation riparian forests (Plummer et al., 1902, Collins and Montgomery, 2001). Hence, an increase in the supply and size of wood delivered to rivers will complicate delineation of the CMZs. In particular, guidelines for the delineation of CMZs (e.g., WFPB, 2001) do not account for the causal link between vertical fluctuations in channel-bed elevation and lateral channel migration in rivers where wood obstructions are common.

We quantified the scale of wood-induced vertical fluctuation in channel-bed and water-surface elevations to evaluate the potential for vertical fluctuation from wood obstructions as an additional mechanism for channel migration. Specifically, we investigated the influence of vertical channel change associated with wood accumulations at eleven locations in Washington state (Fig. 1). First, we use hydraulic modeling of historical wood removal from a natural system to suggest a probable magnitude in vertical fluctuation and then question how widespread the mechanism may be in riparian forests. We then compare model results with vertical fluctuations measured at ten field sites impacted by channel avulsions associated with accumulations of LWD. Finally, we discuss the implications of the magnitude of vertical channel fluctuation for the delineation of the CMZs and the impact on lateral channel migration as maturing riparian forests supply increasing wood volume to rivers.

Section snippets

Hydraulic model

The influence of wood on the morphologic response of rivers can be evaluated by modeling the effects of flow obstructions on water-surface elevations. Conservation of mass for flow through an open channel dictates thatQ=UAwhere Q is discharge, U is mean velocity, and A is the cross sectional area of flow. Resistance to flow is commonly expressed as an empirical function of U, hydraulic radius (R), and water-surface slope (S) using Manning's equation:n=R2/3S1/2Uwhere n is Manning's roughness

Field studies

Motivated by results of the hydraulic modeling, we set out to quantify the vertical change in bed-and water-surface elevations caused by channel-spanning logjams and evaluate their influence on channel migration. We investigated ten additional field sites located in unconfined channels congested with large wood accumulations. With the exception of tributaries surveyed along the West Fork Satsop River, field sites were located in channels with a gradient < 0.02 and watershed sizes ranging from 20

Discussion

Our case studies show that the potential for lateral channel migration is strongly linked to vertical channel adjustments associated with accumulations of wood. Hydraulic modeling of the Ozette system suggests that logjam removal lowered the water-surface elevation by a minimum of 1 m. The hysteresis between river stage and discharge measured on the Deschutes River suggests logjam properties and hydraulic head mutually adjust through changes in jam porosity and water-surface elevation during

Conclusions

Stable logjams mediate the water-surface elevation in unconfined alluvial channels by introducing hydraulic resistance and creating sites of sediment impoundment. Field investigations of the effects of logjams in unconfined rivers in Washington State indicate that aggradation behind logjams can result in vertical fluctuations that exceed the bankfull elevation. Resource management goals to restore riverine habitat by increasing the size and supply of wood to rivers currently deficient in stable

Acknowledgements

We gratefully acknowledge the assistance of C.A. McIntosh with field surveys and the help of Maeve McBride with the HEC-RAS modeling. The manuscript benefited substantially from the comments of two anonymous reviewers.

References (42)

  • R.L. Beschta

    Debris removal and its effects on sedimentation in an Oregon Coast Range stream

    Northwest Science

    (1979)
  • R.E. Bilby

    Role of organic debris dams in regulating the export of dissolved organic and particulate matter from a forested watershed

    Ecology

    (1981)
  • R.E. Bilby et al.

    Changes in characteristics and function of woody debris with increasing size of streams in western Washington

    Transactions of the American Fisheries Society

    (1989)
  • P.A. Bisson et al.

    Large woody debris in forested streams in the Pacific Northwest: past, present, and future

  • M. Boyer et al.

    Guide Technique SDAGE-La Gestion des Boisements de Rivie`re: Dynamique et Fonctions de la Ripisylve

    (1998)
  • C.A. Braudick et al.

    When do logs move in rivers?

    Water Resources Research

    (2000)
  • J.M. Buffington et al.

    Effects of hydraulic roughness on surface textures of gravel-bed rivers

    Water Resources Research

    (1999)
  • H. Chanson

    The Hydraulics of Open Channel Flow

    (1999)
  • B.D. Collins et al.

    Importance of archival and process studies to characterizing pre-settlement riverine geomorphic processes and habitat in the Puget Lowland

  • B.D. Collins et al.

    Forest development, wood jams, and restoration of floodplain rivers in the Puget Lowland

    Restoration Ecology

    (2002)
  • B.D. Collins et al.

    Historic changes in the distribution and functions of large woody debris in Puget Lowland rivers

    Canadian Journal of Fisheries and Aquatic Sciences

    (2002)
  • Cited by (67)

    • Wood in Fluvial Systems

      2022, Treatise on Geomorphology
    • Sediment yield from a forested mountain basin in inland Pacific Northwest: Rates, partitioning, and sources

      2021, Geomorphology
      Citation Excerpt :

      Perhaps more importantly, channel incision, which has been noted in Sullivan Creek (Fig. 8), was likely to take place due to reduced flow resistance. Such incision may be a common consequence of stream clearing in other Pacific Northwest channels (Brummer et al., 2006). Although, based on simple calculations (see Supplemental material), it seems that this source could close the sediment budget, we stop short of attempting to assess its actual contribution.

    • Lotic freshwater: Rivers

      2020, Encyclopedia of the World's Biomes
    View all citing articles on Scopus
    View full text