Channel evolution model

by J. Collison


Channel Evolution Model (CEM) is an approach to explain the complexity of a fluvial system. A fluvial system is constantly changing and evolving, which is the systems attempt to reach equilibrium. A system that is considered stable or in equilibrium is well vegetated, frequently interacts with its floodplain and the sediment is suspended. CEM is used to classify the current stage of the system in order to predict how the system will evolve. Knowing the current stage of a system is incredibly beneficial when alterations to a system are being considered, especially when those alterations are aimed to provide restoration. Schumm and Parker initially posed the idea of CEM (1973).

Stanley A. Schumm wrote The Fluvial System in 1977 in order to make the complex fluvial system more easily understood. Schumm stressed the point of how unstable these systems are and how one needs to understand the upstream and downstream controls that governs the system. Schumm used an empirical approach to demonstrate how complex the fluvial system is. By using the Rainfall-Erosion Facility (REF) at Colorado State University, Schumm was able to model various types of landscapes. The REF utilized 9m x 15m containers of various types of soils that could be altered by changing the slope, intensity of precipitation (from overhead sprinkler system held aloft by a crane), amount of precipitation, location of precipitation and the outlet location (Schumm, 1977).

In one experiment the system was allowed to come to equilibrium (Figure 1A) then the outlet was lowered 10cm. This lowering of the outlet caused incision to occur at the mouth of the system and then propagate upstream (Figure 1B). As the incision moved upstream sediment loads downstream increased greatly and deposition began, which led to a braided channel in the downstream sections of the system (Figure 1C). As the incision decreased upstream the sediment loads downstream decreased, which allowed another incision stage to start. This incision-deposition-incision process would continue until a new equilibrium in the system was reached (Figure 1D) (Schumm and Parker, 1973).

Figure 1: A channels response to the lowering of its base level (Schumm and Parker 1973)

Schumm’s experimental work at REF led to his co-authored book Incised Channels: Morphology, Dynamics and Control (1986). This book focused around the space and time variable aspect of CEM and how to force a system from evolving into later stages. This book is considered “The go to book” for CEM. Simon and Hupp also expanded on the earlier work of Schumm and Parker to create Figure 2 below (1986). This model can be used on any low gradient—less than 2% gradient flow—(eg: braided, wandering, sinuous and anastomosing) channels which may include sections with high gradient—greater than 2% gradient flow—(eg: riffle-pool sequence, rapids, step-pools and cascades) channels. Further information on channel types can be found in Stream Classification Techniques. Figure 2 is a modified version of Simon and Hupp (1986) CEM found in Simon and Rinaldi (2006).

Stages of a Channel

When a channel is disturbed by anthropogenic or natural means the channel will then go through the stages in the above figure in order to reestablish equilibrium.

Stage I – Stable System
This stage is stable system at equilibrium, typically a meandering or straight channel that is well vegetated. The system is in frequent contact with its floodplain. The majority of the sediment is suspended with only a small amount of bedload sediment. This bedload sediment is from a localized aggradation and degradation process called scour. Further information about sediment transport can be found at Sediment Transport Processes. The banks are typically very stable, due to the vegetation and cohesive bank material. Streampoweris steady and constant (Simon and Rinaldi, 2006).

Stage II - Disturbance
Something pushes the system out of equilibrium. A reduction in base height due to channelization is typically the cause of this disturbance. Channelization is the act of removing excess bends and meanders in a system, which creates a system with an increased gradient and more streampower. Although changes in water use (increased base flow) will also push the system out of equilibrium due to the increased stream power. A decrease in base flow can also push the system out of equilibrium. This decrease in flow will reduce the streampower of the system and begin Stage V, aggradation of the system (Simon and Rinaldi, 2006).

Stage III - Degradation
With the increased streampower the system begins to incise in the vicinity of disturbance in the systems attempt to adjust to the lower channel base height. As the disturbed site degrades a knickpoint migrates up channel. Incision at the disturbance continues until the channel reaches its new base height. The knickpoint continues to migrate up channel (Simon and Rinaldi, 2006).

Stage IV – Degradation and Widening
During the down cutting of the channel in Stage III, bank stability decreases as bank height and the steepness of the bank increases. This leads to bank destabilization, mass wasting, of the bank and channel widening. More information about bank stability can be found here, Bank Stability. The degradation of the channel continues to move up channel and some aggradation begins down channel due to a decrease in streampower from the channel reducing its gradient (Simon and Rinaldi, 2006).

Stage V – Aggradation and Widening
Aggradation of the channel is the dominate feature and channel widening continues. Channel degradation moves farther upstream, increasing the sediment load downstream. The lower portions of the channel have a more flattened gradient (due to the initial degradation) and therefore a reduced streampower. This reduction of streampower and high sediment loads from upstream cause the lower portions of the system to begin aggrading. This aggrading migrates upstream like the knickpoint did in Stage III. If the aggrading brings the base height higher than the new base height established in Stage II the system will go through Stage III – Stage V again. This degrading-aggrading-degrading cycle will continue with each cycle becoming less intense until Stage VI is reached (Simon and Rinaldi, 2006).

Stage VI – Quasi Equilibrium
If the system remains disturbance free a new quasi equilibrium will be reached. The old floodplain will form a terrace over the new lowered floodplain. A new floodplain will begin to form in aggraded material that was deposited during Stage IV and V. Over time the establishment of vegetation and meanders in the new floodplain will help to establish stability for the system, creating a new equilibrium (Simon and Rinaldi, 2006).
Figure 2: Channel Evolution Model Stages (Simon and Rinaldi 2006)

Alluvial rivers like the Rio Grande closely followed REM on a large scale in the past before humans started to modify the system. Even with huge structures (Elephant Butte and Cochiti Dam to name two) in place, the Rio Grande still follows the REM but on a much smaller scale and more slowly than in the past.

History of the Middle Rio Grande

The predominate driving force for the Middle Rio Grande (MRG is defined by the reach of the Rio Grande between Elephant Butte Dam and Cochiti Dam) channel evolution, before anthropogenic factors, was the valley uplift that occurs roughly where Rio Salado enters the Rio Grande (Massong et al., 2010). This uplift (Stage II) would increase the channel gradient downstream of the uplift, which would trigger Stage III of the CEM. The channel degradation would migrate upstream and eventually up tributaries and arroyos. The Rio Grande would then evolve through Stages IV, V and VI at which point if the uplift remained inactive the system would reestablish equilibrium. This new equilibrium would remain until another uplift occurred, which would lead to another Stage III - Stage VI cycle. This degradation/aggradation cycle has been happening for the last 3000 years on the Rio Grande (Klinger and Klawon, 2003). The last degradation cycle ended in the 1950’s (Massong et al., 2010).

Anthropogenic forces at Large

Historically, the Rio Grande has been prone to massive floods, with one of the largest and longest floods happening in 1941 and a major flood the following year. The major flood in 1942 led to several flood control acts being passed by Congress, with Cochiti Dam being the largest flood control act. Since the completion of Cochiti Dam in 1973 the MRG has begun adapting to its new constraints (Cochiti Dam and Elephant Butte). The section of the MRG upriver of Rio Salado is responding by entering stage III (Belen, NM), due to decrease of sediment, and the downstream section of Rio Salado entering stage VI, with some areas (Rail Road Bridge at San Marcial, NM) becoming so heavily aggraded that the river’s base level is higher than the surrounding floodplains.

Case Studies

The first case study was done by Massong et al. (2010). This case study focused on an area near Belen, NM. Through a series of aerial pictures of roughly the same location (Figure 3) the gradual channelization of the Rio Grande can be seen. Figure 3A shows a heavily braided system with many bars and islands forming, which would be considered Stage IV/Stage V. Figure 3B shows a slightly less braided system with vegetation starting to form on the larger bars/islands. This system would be considered to be entering Stage VI. Three years later in Figure 3C the majority of the flow is located within the main channel. The main channel looks to be forming some meanders and straight sections. This system is still working towards Stage VI. Finally in Figure 3D the system is firmly channelized with clear meanders and straight sections formed. The vegetation on the bars/islands is well established. This system is considered to be at quasi equilibrium, Stage VI, but not fully at equilibrium. The meanders appear to be doing some eroding at the cut banks and depositing at the point bars. Figure 3 is a great example of the mid-later stages of a CEM.

Figure 3: Near Belen, NM. Channelization of the Rio Grande. A:2000, B:2002, C:2005, D:2006 (Massong et al., 2010)

Another case study done by Massong et al., shows the effects of a heavily aggraded section of the Rio Grande near the railroad bridge at San Marcial, Figure 4 (2010). This section of the MRG has historically been prone to heavy aggradation. A flood in 1929 deposited more than 7 feet of sediment in the town of San Marcial and then in 1937 another flood destroyed any remaining structures in San Marcial (Van Citters, 2000). In recent times (1991, 1995, 2005 and 2009), the Rio Grande has become completely plugged in the vicinity of the railroad bridge. Early speculations blamed Elephant Butte for the high aggradation in this area, but recent papers (Boroughs 2005, Lai 2009 and Van Citters 2000) point to the channel’s characteristics as being the cause of the high aggradation, Stage V CEM.


Figure 4: Rio Grande during the sediment plug of 2005, near Railroad Bridge at San Marcial. (Massong et al., 2010)

Channel Evolution Model in Relation to River Restoration

Restoration can be broken down into two categories, passive or active. Passive restoration focuses on correcting the cause of the degradation of the system and then allowing the system to reestablish equilibrium without any more human intervention. Active restoration uses a more engineered approach to force the system to the desired result. Active restoration typically requires more maintenance then passive restoration (Tetra Tech EM Inc. 2004). Further information about restoration goals can be found at, Setting Restoration Goals and Objectives.

Fully understanding the current CEM stage of the section of river that’s going to be affected by a restoration project can greatly increase the success of the project. Also, knowing the controls that govern the system upstream and downstream can increase the longevity of the restoration project. These governing controls can alter the stage of the restored section of the system in the future.

River restoration must use different restoration approaches for different CEM stages. A great example of a restoration project using different approaches for the different stages of a CEM is the Las Vegas Wash Capital Improvement Plan. This project over the years (1999-2012+) has incorporated various forms of channel bed stabilization, channel bank protection, flood wave energy dissipation structures, and wetlands and riparian revegetation. This project has incorporated detailed planning and investigation at every step of the project in order to restore their system in the best possible way. Most restoration projects do not have the luxury of funds this project has had, over $100 million in funds to date, their investigation and design process can be scaled down to almost any size restoration project. Based on a systems CEM the following guide lines for restoration should be considered.

Stage I – Stable System
If the system is stable, chances are restoration is not needed, although a stable system does not always correlate to a healthy system. If restoration is needed to promote a healthy system, passive restoration should be used to correct the problem that is causing the need for restoration. Active restoration should be avoided on a stable system if at all possible. If the active restoration is too extreme it might push the system out of equilibrium, which will undo any positive benefit the restoration project gained.

Stage II – Disturbance
If anthropogenic system instability is foreseeable, active and passive restoration process should be used in order to keep the system in equilibrium. If the system is already unstable, steps to stabilize the system should be used. Active restoration such as bed stabilization should be used to prevent Stage III to forming. Depending on the cause, a reduction in streampower can help to prevent incision. Disturbances do not always cause the system to enter Stage III, they can also cause the system to enter Stage V, aggradation. A reduction in flow can led to reduced stream power and an increase of deposition within the system. Constraining the flow within the system can help increase streampower and reduce aggradation.

Stage III - Degradation
If incision has already begun, attempt to stabilize the channel bed using active restoration in order to control the degradation. All effort should be spent to prevent further degradation of the channel and prevent the knickpoint from migrating farther up the system. If the knickpoint has moved outside the scope of the project, effort should be spent on increasing the streampower within the project area to prevent future aggradation from occurring. Passive restoration should be avoided until the system is more equilibrated.

Stage IV – Degradation and Widening
Depending on the degree of degradation passive restoration could be used to help stabilize the channel bed. Passive restoration can also be used to help stabilize the banks and prevent widening. Active restoration would need to be used to stabilize the channel bed if the streampower is too intense for passive means. If mass wasting and bank widening are too extreme, active restoration will need to be used to force the system into a more equilibrated state.

Stage V – Aggradation and Widening
If aggradation within the system is very active, steps should be taken to remove the excess deposition and stabilize the system before restoration should take place. If the system has aggraded far above “normal” levels, dredging of the channel might have to take place. Another less aggressive way to remove excess sedimentation is to constrict the channel in order to increase streampower and allow the system to naturally remove the sediment. If the constriction method is used, implications of higher sediment loads on downstream sections of the system should be considered. Bank widening will be less severe than in Stage IV, which should allow for passive restoration to stabilize the system.

Stage VI – Quasi Equilibrium
If the system has reached a state of quasi equilibrium active and/or passive restoration should be used to prevent the system from entering Stage III at all costs. All restoration efforts should focus on stabilizing the system in order to create a lasting equilibrium state. System governing factors upstream and downstream that might push the system out of equilibrium should be considered in designing the restoration project in order to account for future disruptions within the system.


Knowing the correct CEM stage of a system can mean the difference between a successful and unsuccessful restoration project. CEM and other Stream Classification Techniques should be used to help determine the direction and goals of restoration. When designing a restoration project, upstream and downstream stages should be considered in order for the project to have a lasting effect. Another key to a successful restoration project is to design/build a system with the controlling factors of the system (dams, lakes, tributaries, changing water usage, etc.) in mind. Detailed Hydrology and Fluvial Geomorphology studies should not be skimped on in order to save a few “bucks”. Overall the most important aspect when planning a restoration project is knowing how the system is currently acting and how restoration might alter the system.


Boroughs, C. B., 2005. Sediment Plug Computer Modeling Study, Tiffany Junction Reach, Middle Rio Grande Project, Upper Colorado Region. U.S. Department of the Interior, Bureau of Reclamation, Albuquerque Area Office. Technical Report, Oct 2005, p. 122.
Klinger, R.E., and Klawon, J.E., 2003. Late Holocene Alluvial Stratigraphy and Geomorphology of the Little Colorado River Between Holbrook and Winslow, Arizona, Geological Society of America Abstracts with Programs, Paper No. 26-7.
Lai, Yong, 2009. Sediment Plug Prediction on the Rio Grande with SRH-2D Model. Sedimentation and River Hydraulics Group, Technical Service Center, Bureau of Reclamation, United States Department of the Interior, Denver CO.
Massong, T., Makar P., Bauer T., 2010. Planform Evolution Model for the Middle Rio Grande, NM. 2nd Joint Federal Interagency Conference, las Vegas, NV, June 27 – July 1.
Schumm, A. Stanley, 1977. The Fluvial System. The Blackburn Press.
Schumm, A. and Parker, R. S., 1973, Implications of complex response of drainage systems for quaternary alluvial stratigraphy: Nature (Physical Science) v. 243, p. 99-100.
Schumm, A., Watson, C., and Harvey, M., 1986. Incised Channels, Morphology, Dynamics and Controls, Water Resources Publications, Littelton, CO. 200pp
Simon, A., Rinaldi M., 2006. Disturbance, stream incision, and channel evolution: The roles of excess transport capacity and boundary materials in controlling channel response. Geomorphology 79 (2006) 361-383.
Simon, A., Hupp, C.R., 1986. Channel evolution in modified Tennessee channels. Proceedings, Fourth Federal Interagency Sedimentation Conference, Las Vegas,March 24–27, 1986, vol. 2, pp. 5–71–5–82.
Tetra Tech EM Inc., 2004. Habitat Restoration Plan for the Middle Rio Grande. Prepared for: Middle Rio Grande Endangered Species Act Collaborative Program Habitat Restoration Subcommittee, September 2004.
Van Citters, K., 2000. Historic Engineering Overview of the San Marcial Railroad Bridge. Submitted to U.S. Army Corps of Engineers, by Office of Contract Archeology, University of New Mexico, OCA/UNM Report No. 185-665, 63.