Causes of Streambank Erosion
Shortcut URL: https://serc.carleton.edu/43576
Location
Continent: North America
Country: United States of America
State/Province:
City/Town:
UTM coordinates and datum: none
Setting
Climate Setting: Humid
Tectonic setting: Intracratonic Basin
Type: Process
Show caption
Hide
Conceptual drawing of rotational wedge failure. Applicable to cohesive material.
Details
Description
Streambank erosion is a natural phenomenon but can become accelerated because of changes at the watershed and/or reach level. Background conditions are established using old photographs and maps and compare rates over time, in the context of floods, and watershed and reach level changes. These are combined with recognizing reach level changes in stage of channel evolution, scour patterns, sediment deposits, side slopes, slumping, and cultural features. Changes at the watershed level might be caused by logging, mining, road building, Changes at the reach level might be riparian area modifications due to vegetation change or loss, animal use, stream channel straightening, or channel excavation and/or removal of material (stream bed mining), avulsion (abrupt change in stream course, such as meander neck cut-off) or climate change.
Investigate whether bank instability is caused by cantilever failure (Figures 1 and 2); planer failure (Figure 3); rotational failure (Figure 4); preferential flow failure including piping (Hagerty, 1991); high pore pressure; liquefaction and seepage forces especially during the falling stage of floods; popout failure; or because bank height exceeds some critical bank height (Schumm et al. 1984, Reckendorf and Tice, 2001, 2002, Reckendorf, 1996) (Figure 5). Investigate if failure mechanisms are accelerated or modified by a stratigraphy caused cantilever (Reckendorf, 2009, 2008, 2001, 1996, 1989). Is failure mechanism accelerated by: Rc/Wbkf geometry as discussed in, Bagnold (1960), Thorne et al. (1997) Welch and Wright (2005), Southerland and Reckendorf (2008); flow condition (peak, duration, angle of attack parallel to perpendicular to angle of failure), helicoidal flow, (Wolman, 1959, Hooke, 1979, Thorne et al. 1997, Welch and Wright, 2005); pre-wetting (Knighton, 1995, and Thorne et al. (1997); boundary shear stress; depth of bed scour along eroding bank; local sedimentation (Reckendorf, 2009, 2008), and Rosgen 2006), root density and depth (Reckendorf, 2001, and Rosgen, 2006), waves (Reckendorf, 1989); ice condition (freeze thaw, Knighton, 1995, and gouging); desiccation (Thorne et al. 1997); animal burrows (Reckendorf, 2008), or large woody debris jams causing avulsion.
Understanding historic channel changes will consistently help us understand streambank erosion The Channel Evolution Model (CEM) of Schumm et al. (1984), and Simon (1989) provide an excellent perspective. The CEM has been used to field identity the CEM stages on dozens of streams in dozens of states (Reckendorf and Steffan, 2006 ). The CEM Stage II stage of downcutting is often a result of channel straightening (Reckendorf and Tice, 2003). This creates a very high bank relative to a critical bank height where the streambank fails (Figure 5) , and a headcut that migrates upstream. Stage III in the CEM, reflects widening. Streambank protection work, drop structures, soil bioengineering, or stream restoration work should not proceed without identifying the stage of channel evolution, with the associated headcut migration.
Accelerated streambank erosion at the reach level is often modified by local sedimentation patterns that redirect flow, to the extent that avulsions occur where not expected. Landslides and debris flow often provide such a large quantity of coarse material to streams that the lateral shear accelerates streambank erosion, instead of transporting the coarse material (Reckendorf, 2006, 2008). Debris flows tend to migrate downstream in pulses (Benda and Dunne, 1997a, 1997b) and Reckendorf (1996, 2006, 2008, ), reworking the course material during high flow. Often streams don't have the competence to transport the commonly available largest sediment particles resulting in lag deposits that accumulates. This coarse accumulation exacerbate the bank erosion, as the river flanks the coarse deposit.
Watershed level changes like logging can cause increased sedimentation downstream that completely bury the natural channel and adjacent flood plain. This results in a whole new cycle of streambank erosion (Trimble, 1982).
There are hundreds of miles of streambank, where the author has concluded that stratigraphy (especially washing fines and sand matrix out of gravel strata) is the principal cause of cantilever failure, along gravel bed river streambanks. For example in just one county (Clackamas County in Oregon) five rivers (Sandy, Salmon, Zigzag, Clackamas, Molalla, representing hundred miles of river, streambank have primarily cantilever failure that was caused by high water removal of sands and fines matrix material developing a cantilever conditions for failure. In one study by the author on over 150 miles of streambanks in the Walla Walla and Columbia Counties in WA (Reckendorf and Tice , 2001, 2003), the majority of the streambank failures were from downcutting and widening associated with cantilever failure, because of the removal of fines and sand matrix material from the gravel strata.
In a post project appraisal study of the failure of engineering log jams in Washington (Southerland and Reckendorf, 2008) streams with a Rc/WBkf of less than 2.5 were found to be the most prone to failure along meandering streams. These type of tight meander curves are thought to have deep thalweg along the outside curves and high helicoidal flow contributing to streambank failure (Welch and Wright 2005).
Understanding the geomorphic history of a stream helps in our understanding of the nature of stratigraphy. The geomorphic traditional view that meandering streams are composed of meander vertical accretion deposits overlaid by vertical accretion deposits, may have distorted our perceptions of the materials in the gravel stratigraphic deposits. On occasions studies such as Karlstrom (1964), on Kenai River, AL have shown that the gravel stratigraphy was formed by a braided stream with a paleo-channel and paleo-hydrology, and that the existing stream is underfit. The Kenai River, AL, paleo-channel formed a coarse cobble gravel strata, that was later overlain by vertical accretion deposits. These Pleistocene and Holocene braided stream gravel deposits are now apparent in flood plain and low terrace streambanks, and are presently extensively undercut by cantilever failure. High flows and waves impacts now remove sand and fines matrix material from the cobble strata (Reckendorf and Seale, 1991, and Reckendorf, 1989). Over time, with a reduced runoff in the Holocene and reduced base level, streams down-cut through the prior braided gravel deposits and formed single or multiple-thread meandering streams, with the vulnerable coarse cobble gravel stratigraphy in the streambanks. It is not uncommon to find a coarse cobble stratigraphy in a streambanks with cantilever failure and that also show a bimodal distribution of sediment on meandering streams point bars verses the streambed. The bimodal bar stratigraphy are often the result of avulsion (Reckendorf 2010), and once overlain by vertical accretion deposits the history of the stratigraphy is difficult to interpret.
Investigate whether bank instability is caused by cantilever failure (Figures 1 and 2); planer failure (Figure 3); rotational failure (Figure 4); preferential flow failure including piping (Hagerty, 1991); high pore pressure; liquefaction and seepage forces especially during the falling stage of floods; popout failure; or because bank height exceeds some critical bank height (Schumm et al. 1984, Reckendorf and Tice, 2001, 2002, Reckendorf, 1996) (Figure 5). Investigate if failure mechanisms are accelerated or modified by a stratigraphy caused cantilever (Reckendorf, 2009, 2008, 2001, 1996, 1989). Is failure mechanism accelerated by: Rc/Wbkf geometry as discussed in, Bagnold (1960), Thorne et al. (1997) Welch and Wright (2005), Southerland and Reckendorf (2008); flow condition (peak, duration, angle of attack parallel to perpendicular to angle of failure), helicoidal flow, (Wolman, 1959, Hooke, 1979, Thorne et al. 1997, Welch and Wright, 2005); pre-wetting (Knighton, 1995, and Thorne et al. (1997); boundary shear stress; depth of bed scour along eroding bank; local sedimentation (Reckendorf, 2009, 2008), and Rosgen 2006), root density and depth (Reckendorf, 2001, and Rosgen, 2006), waves (Reckendorf, 1989); ice condition (freeze thaw, Knighton, 1995, and gouging); desiccation (Thorne et al. 1997); animal burrows (Reckendorf, 2008), or large woody debris jams causing avulsion.
Understanding historic channel changes will consistently help us understand streambank erosion The Channel Evolution Model (CEM) of Schumm et al. (1984), and Simon (1989) provide an excellent perspective. The CEM has been used to field identity the CEM stages on dozens of streams in dozens of states (Reckendorf and Steffan, 2006 ). The CEM Stage II stage of downcutting is often a result of channel straightening (Reckendorf and Tice, 2003). This creates a very high bank relative to a critical bank height where the streambank fails (Figure 5) , and a headcut that migrates upstream. Stage III in the CEM, reflects widening. Streambank protection work, drop structures, soil bioengineering, or stream restoration work should not proceed without identifying the stage of channel evolution, with the associated headcut migration.
Accelerated streambank erosion at the reach level is often modified by local sedimentation patterns that redirect flow, to the extent that avulsions occur where not expected. Landslides and debris flow often provide such a large quantity of coarse material to streams that the lateral shear accelerates streambank erosion, instead of transporting the coarse material (Reckendorf, 2006, 2008). Debris flows tend to migrate downstream in pulses (Benda and Dunne, 1997a, 1997b) and Reckendorf (1996, 2006, 2008, ), reworking the course material during high flow. Often streams don't have the competence to transport the commonly available largest sediment particles resulting in lag deposits that accumulates. This coarse accumulation exacerbate the bank erosion, as the river flanks the coarse deposit.
Watershed level changes like logging can cause increased sedimentation downstream that completely bury the natural channel and adjacent flood plain. This results in a whole new cycle of streambank erosion (Trimble, 1982).
There are hundreds of miles of streambank, where the author has concluded that stratigraphy (especially washing fines and sand matrix out of gravel strata) is the principal cause of cantilever failure, along gravel bed river streambanks. For example in just one county (Clackamas County in Oregon) five rivers (Sandy, Salmon, Zigzag, Clackamas, Molalla, representing hundred miles of river, streambank have primarily cantilever failure that was caused by high water removal of sands and fines matrix material developing a cantilever conditions for failure. In one study by the author on over 150 miles of streambanks in the Walla Walla and Columbia Counties in WA (Reckendorf and Tice , 2001, 2003), the majority of the streambank failures were from downcutting and widening associated with cantilever failure, because of the removal of fines and sand matrix material from the gravel strata.
In a post project appraisal study of the failure of engineering log jams in Washington (Southerland and Reckendorf, 2008) streams with a Rc/WBkf of less than 2.5 were found to be the most prone to failure along meandering streams. These type of tight meander curves are thought to have deep thalweg along the outside curves and high helicoidal flow contributing to streambank failure (Welch and Wright 2005).
Understanding the geomorphic history of a stream helps in our understanding of the nature of stratigraphy. The geomorphic traditional view that meandering streams are composed of meander vertical accretion deposits overlaid by vertical accretion deposits, may have distorted our perceptions of the materials in the gravel stratigraphic deposits. On occasions studies such as Karlstrom (1964), on Kenai River, AL have shown that the gravel stratigraphy was formed by a braided stream with a paleo-channel and paleo-hydrology, and that the existing stream is underfit. The Kenai River, AL, paleo-channel formed a coarse cobble gravel strata, that was later overlain by vertical accretion deposits. These Pleistocene and Holocene braided stream gravel deposits are now apparent in flood plain and low terrace streambanks, and are presently extensively undercut by cantilever failure. High flows and waves impacts now remove sand and fines matrix material from the cobble strata (Reckendorf and Seale, 1991, and Reckendorf, 1989). Over time, with a reduced runoff in the Holocene and reduced base level, streams down-cut through the prior braided gravel deposits and formed single or multiple-thread meandering streams, with the vulnerable coarse cobble gravel stratigraphy in the streambanks. It is not uncommon to find a coarse cobble stratigraphy in a streambanks with cantilever failure and that also show a bimodal distribution of sediment on meandering streams point bars verses the streambed. The bimodal bar stratigraphy are often the result of avulsion (Reckendorf 2010), and once overlain by vertical accretion deposits the history of the stratigraphy is difficult to interpret.
Associated References
- Bagnold, R. 1960. Some Aspectcs of the Shape of River Meanders. USGS Professional Paper 282-A.
- Benda, L. and Dunne, L. 1997a. Stochastic forcing of sediment supply to the channel network from landsliding and debris flows. Water Resources research 33: 2849-2836.
- Benda, L. and Dunne, L. 1997b, Stochastic forcing of sediment storage routing and storage. Water Resources Research 33: 1865-2880.
- Hagerty, D. 1991. Piping/sapping erosion. basic considerations, Proceedings of the American Society of Civil Engineers, Journal of Hydraulics Engineering, 117: 991-1008.
- Hooke, J. 1979. An Analysis of the processes of riverbank erosion. Journal of Hydrology, 42: 39-62.
- Knighton, D. 1995. Fluvial Processes and Forms. Edward Arnold, New York 218pp.
- Reckendorf, F. 2010. East Fork Lewis River, including West Daybreak Park project, Fluvial geomorphology and erosion and sediment evaluation. Reckendorf and Associates, Salem, OR 42p.
- Reckendorf, F. 2009. Causes of streambank erosion. Abstract 159236, Geological Society of America Annual Meeting, From Volcanoes to Vinyards, Living with Dynamic Landscapes. October 18 - 21, 2009. Portland, OR.
- Reckendorf, F. 2008. Bar Scalping of Donaldson, Barker, Dill, Gomes, and Waldron Gravel Bars in Tillamook County, Report prepared for Coast Wide Readymix, Tillamook, OR. Reckendorf and Associates, Salem, OR 77p.
- Reckendorf, F. 2006. Environmentally sensitive gravel bar scalpking. Federal Interagency Sedimentation Conference, April 2-6, 2006. Session 4B-3. CD pdf.
- Reckendorf, F. and Steffen, L. 2006. Regional application of stream systems in planning and design of stream stabilization projects. ASCE, World Environmental and Water Resources Congress, Omaha, NB CD, pdf.
- Reckendorf, F. and Tice, B. 2003. Basin wide stream degradation limits spawning and rearing habitat for salmonids. GSA Conference Horizons, Seattle 2003. Abstract 8-16, p25.
- Reckendorf, F. and Tice, B.. 2001. Rapid assessment procedure for aquatic habitat, riparian and streambank (RAPFAHRS), Proceedings of the Seventh Federal Interagency Sedimentation conference, March 25-29, 2001, Reno Nevada. 7p.
- Reckendorf, F. 1996. Stream system evaluations with interpretations for habitat rehabilitation, restoration, and land use planning. Conference Dinner Presentation, Sixth Federal Interagency Sedimentation conference, Sahara Hotel, Las Vegas, Nevada, March 12, 1996. Reckendorf and Associates, Salem OR. 23p.
- Reckendorf, F. and Saele, L. 1991. Kenai River, Bank Inventory Report. USDA, Soil Conservation Service. West Technical Service, July,1991. 92p.
- Reckendorf, F. 1989. Kenai River Streambank Erosion Special Report. USDA, Soil Conservation Service, West National Technical Center, Portland, Oregon, December 1989. 55p.
- Rosgen, D. 2006. Watershed Assessment of River Stability and Sediment Supply (WARSSS). Wildland Hydrology, Lakewood, Colorado. 533p.
- Schumm, S. Harvey, M. , and Watson, C. 1984. Incised Channel, Morphology, Dynamics, and Control. Water Resources Publications, Littleton, CO. 199p.
- Simon, A. 1989. A model of channel response in distributed alluvial channels. Earth Processes and Landforms 14(1)11-26.
- Thorne, C., Hey, R., Newson, M. 1997. Applied Fluvial Geomorphology in River Engineering and Management. John Wiley and Sons, New York. 376p.
- Wollman, G.. 1959. Factors nfluencing erosion of a cohesive river bank. American Journal of Science, 257, 204-216.
- Trimble, S. and Lund, S. 1982. Soil Conservation and Reduction of Erosion and Sedimentation in the Coon Creek, Basin. Wisconsin. USGS Professional Paper 1234.
- Welch, S. and Wright, S. 2005. Design of Stream Barbs. USDA Natural Resources Conservation Service, Technical Note 23. 23p. asce.org/conference/ewri2006/
- www.gcdamp.gov/aboutamp/pfs/sed-transport-rpt96-04pdf