Stream Restoration Projects
Incorporating Channel Degradation and Restoration into Watershed Models
Introduction
 |
|
Sediment plume in Stroubles Creek, Blacksburg, Virginia, USA. |
Excessive channel degradation reduces water quality through increased turbidity and the transport of sediment-bound pollutants. Sediment is the fourth leading cause of water quality impairment nationwide (EPA, 2005).
Watershed modeling software is essential in the development of TMDL studies. The software helps researchers simplify complex watershed systems and allows them to determine the sources of impairment and the reductions required to meet water quality standards.
Sponsored by the US Environmental Protection Agency (EPA), the goal of this project was to summarize existing models and software that evaluate sediment contributions to streams due to channel degradation from streambed scour and streambank retreat. Additionally, a case study was conducted to compare three software packages with a range in model complexity (GWLF, SWAT, and CONCEPTS) to field measurements of channel degradation.
© VT-BSE-TMDL Center 2006 | Updated 1 November 2006
Funded by US EPA Grant AW-83238501 Top
Streambank Retreat
 |
|
Monitoring by the USDA-ARS along Goodwin Creek, Mississippi, USA.
|
- subaerial processess and erosion;
- fluvial erosion; and
- bank failure.
Subaerial processes are climate-related phenomena that reduce soil strength, inducing direct erosion and making the bank more susceptible to fluvial erosion (e.g. frost heave, desiccation cracking). Fluvial erosion is the direct removal of soil particles or aggregates from the stream bed or bank toe by stream flow (hereafter called streambank erosion), while the collapse of streambanks due to slope instability is referred to as bank failure or mass wasting.
Streambank retreat is a cyclic process where subaerial processes weaken the soil surface, causing subaerial erosion. During storm events, soil is eroded from the bank toe, leaving an unstable bank. Following a drop in water level, the streambank collapses, depositing additional material at the bank toe.
© VT-BSE-TMDL Center 2006 | Updated 28 August 2006
Funded by US EPA Grant AW-83238501 Top
Equations
Sediment transport in stream channels is difficult to calculate. Due to the complexity of in-stream sediment transport processes no single equation has yet been developed to simulate the movement of particles from all size classes. The sediment transport capacity of individual size classes must be analyzed separately. This requirement demands watershed and stream modeling software employ a suite of transport equations to simulate the movement of all sediment within the basin. Listed below are some of the most common sediment transport equations used in watershed and river modeling software:
- Einstein’s Deposition Model (1968)
- Duboys (1879)
- Ackers and White (1973)
- Meyer-Peter and Müller (1948)
- Laursen (1958)
- Toffaleti (1969)
- Engelund and Hansen (1972)
- Brownlie (1981)
- Yang Sand (1973) and Gravel (1984)
Models
As the importance of channel degradation to stream sediment loadings has been recognized, watershed and stream model developers have incorporated channel degradation subroutines into existing model structures. Both watershed- and reach-scale models are available with varying levels of complexity. Most models only represent fluvial erosion, calculating sediment contributions from the channel boundary based on a critical shear stress and stream sediment transport capacity. More complex models, such as CCHE-1-D, CONCEPTS, and GSTAR-1D, represent both fluvial erosion and mass wasting. Screening-level models include the watershed models AnnAGNPS, ANSWERS-2000, GWLF, HEC-6, HSPF, and WARMF and the reach-scale model SAM. Planning-level models include the watershed model SWAT 2000 and the channel models CCHE-1D and GSTAR 1-D. CONCEPTS, a reach-scale model developed by the US Department of Agriculture's Agricultural Research Service, was specifically designed to model changes in channel morphology. This research-level model permits assessment of stream restoration practices, including the influence of riparian vegetation on bank stability.
When modeling the impact of stream restoration designs, channel degradation process simulation is necessary. Simple empirical models may provide reasonable preliminary estimates of channel degradation sediment loadings, but do not afford the parameter flexibility required to model sediment movement at the reach level. Reach-scale models are required to adequately simulate channel level changes associated with stream restoration practices. Increased channel-level detail available in reach-scale models allows the user to simulate changes in riparian areas through the adjustment of multiple cross section roughness values. Detailed soil strength parameters can also be manipulated to simulate changes in root reinforcement. These detailed channel-level parameters make reach-scale models the only suitable choice if before and after affects of stream restoration are the goal of a modeling effort.
More detailed information on these models can be found in the project’s final report available for download.
For this study eight models listed below were reviewed. However, only GWLF, SWAT and CONCEPTS were evaluated as part of the case study.
AnnAGNPS
AnnAGNPS is a continuous-simulation, multi-event modification of single-event model AGNPS with improved technology and the addition of new features. The model can be used to predict non-point source pollutant loadings from agricultural watersheds. It is a tool for comparing the effects of implementing various conservation alternatives within the watershed. Cropping systems, fertilizer application rates, water and dissolved nutrients from point sources, sediment with attached chemicals from gullies, soluble nutrient contributions from feedlots, and the effect of terraced fields can be modeled (http://www3.bae.ncsu.edu/Regional-Bulletins/Modeling-Bulletin/bosch-annagnps-bulletin-manuscript.html, access November 15, 2006).
AnnAGNPS simulated fluvial erosioin using the Bagnold stream power channel degradation algorithm.
Answers2000
ANSWERS-2000 is a distributed parameter, physically-based, continuous simulation, farm or watershed scale, upland planning model developed for evaluating the effectiveness of agricultural and urban BMPs in reducing sediment and nutrient delivery to streams in surface runoff and leaching of nitrogen through the root zone. The model is intended for use by planners on ungaged watersheds where data for model calibration is not available (http://www3.bae.ncsu.edu/Regional-Bulletins/Modeling-Bulletin/answers2k-draft0.html, accessed November 15, 2006).
Answers2000 simulates fluvial erosion using critical shear stress and transport capacity channel degradation algorithms. Channel adjustment is considered for widenin after nonerodible bed layer is reached.
CCHE-1D
CCHE-1D is a one-dimensional flow and sediment transport model for channel networks. Its hydrodynamic module computes unsteady flows in channels of compound cross sections, accounting for the effects of in-stream hydraulic structures. The sediment transport module computes non-equilibrium transport of non-uniform sediment mixtures. (http://www.ncche.olemiss.edu/content/software/cche1d/applications.pdf, accessed November 15, 2006)
CCHE-1D simulates fluvial erosion and planar bank failure. Channel degradation is accounted for using the critical shear stream algorithm. Channel adjustment is considered by adjusting depth and width based on erosion rate predictions and is designed to assess scour and deposition at hydraulic structures.
CONCEPTS
The Conservation Channel Evolution and Pollutant Transport System (CONCEPTS) is a computer model created by the U.S. Department of Agriculture (USDA), Agricultural Research Service (ARS) at the National Sedimentation Laboratory (NSL) in Oxford, Mississippi. Two versions of the CONCEPTS model were created to simulate open-channel flow hydraulics, sediment transport, and channel morphology (Langendoen, 2000). A watershed-scale version was created to simulate watershed-scale processes and the connectivity of stream networks. The stream corridor version is a reach-scale model created to focus on the hydraulics, sediment movement, and channel shaping processes with increased detail in modeling in-stream processes. The CONCEPTS model was designed as a tool for the assessment of stream corridor restoration projects. When combined with watershed-scale modeling programs, CONCEPTS may be used to assess the long term effectiveness of restoration efforts and provide engineers, planners, and ecologists with quantitative simulation output useful in design implementation procedures.
CONCEPTS simulates fluvial erosion and planar bank failure. Channel degradation is accounted for using the critical shear and transport capacity algorithms. Channel adjustment is considered by width and depth.
Langendoen, E. J. 2000. CONCEPTS – Conservational Channel Evolution and Pollutant Transport System. United States Department of Agriculture – Agricultural Research Service National Sedimentation Laboratory. Oxford, MS.
GSTAR-1D
GSTAR-1D (Generalized Sediment Transport for Alluvial Rivers – One Dimension) is a hydraulic and sediment transport numerical model developed to simulate flows in rivers and channels with or without movable boundaries (http://www.usbr.gov/pmts/sediment/model/gstar1d/index.html , accessed November 15, 2006).
GSTAR-1D simulates fluvial erosion and planar bank failure. Channel degradation is accounted for using critical shear stress and angle of repose adjustment. Width and depth channel adjustment is considered.
GWLF
Researchers at Cornell University developed the Generalized Watershed Loading Function (GWLF) model for predicting sediment loadings to ungaged streams. The program was designed to use readily available data for the prediction of sediment and nutrient movement over time periods ranging from a few months to years. GWLF is suitable for the simulation of small homogeneous watersheds, but may be applied to larger watersheds on a subwatershed by subwatershed basis.
GWLF using an empirical approach for channel degradation as a function of monthly flow volume, % development, animal density, curve number, and channel slope.
HEC-6
HEC-6 is a one-dimensional sediment transport model that calculates water surface and sediment bed surface profiles by computing the interaction between sediment material in the streambed and the flowing water-sediment mixture. The total sediment load is computed for each cross section along with the trap efficiencies for clays, silts, and sands. The change in bed elevation, water surface elevation, and thalweg elevation are also computed for each cross section. Dredging can be simulated and reservoir deposition can be analyzed with the model http://www.hec.usace.army.mil/software/legacysoftware/hec6/hec6.htm, accessed November 15, 2006).
HEC-6 simulates fluvial erosion using the Exner equation, bed sorting, critical shear stress and transport capacity for channel degradation. Vertical adjustment for user defined erodible channel is considered.
HSPF
The HSPF model simulates nonpoint source runoff and pollutant loadings, performs flow routing through streams, and simulates in-stream water quality processes. HSPF estimates runoff from both pervious and impervious parts of the watershed and stream flow in the channel network.
HSPF simulates fluvial erosion using critical shear stress, Parthenaiades’ equation, and Krone’s formula.
SAM
SAM is an integrated system of computer programs developed under the Flood Damage Reduction and Stream Restoration Research Program sponsored by the U.S. Army Corps of Engineers. It is intended to be used primarily as an aid in the design of stable channels. The SAM package enables the user to evaluate the hydraulics, sediment transport, and sediment yield for representative stream cross sections. The programs are not considered to be a model in the sense of evaluating the hydraulics and sediment transport characteristics of an entire stream reach. The sediment transport algorithms in SAM do not compute bed elevation change (erosion and deposition), only sediment transport capacity based on computed hydraulics (http://chl.erdc.usace.army.mil/library/publications/chetn/pdf/chetn-vii-7.pdf, accessed November 15, 2006).
SAM accounts for fluvial erosion using stream power and bed material transport equations.
SWAT2000
The Soil and Water Assessment Tool (SWAT) was developed by Dr. Jeff Arnold of the Grassland, Soil, and Water Research Laboratory in Temple, Texas for the USDA Agricultural Research Service (Neitsch et al., 2002). SWAT is a physically based watershed-scale model created to predict impacts of “land management practices on water, sediment, and agricultural chemical yields in large complex watersheds with varying soils, land uses, and management conditions over long periods of time.
SWAT2000 simulates fluvial erosion using the Bagnaold stream power channel degradation algorithm.
WARMF
To facilitate TMDL analysis and watershed planning, WARMF was developed under sponsorship from the Electric Power Research Institute (EPRI) as a decision support system for watershed management. The system provides a road map to calculate TMDLs for most conventional pollutants (coliform, TSS, BOD, nutrients). It also provides a road map to guide stakeholders to reach consensus on an implementation plan. The Engineering Module is a GIS-based watershed model that calculates daily runoff, shallow ground water flow, hydrology and water quality of a river basin (http://www.epa.gov/ATHENS/wwqtsc/html/warmf.html, accessed November 15, 2006).
WARMF simulates fluvial erosion using critical velocity and transport capacity algorithms.
Case Study
Introduction
A case study was conducted within the Stroubles Creek watershed, in the Town of Blacksburg, Virginia, USA, to compare GWLF, SWAT, and CONCEPTS channel degradation algorithms. An erosion pin monitoring study provided observed data for comparison with model predictions.
The Stroubles Creek watershed, located in Montgomery County, Virginia, USA, is approximately 2,500 ha of urban, residential, agricultural, and forested land, which includes the Town of Blacksburg and the Virginia Tech main campus (see figure below). Urban land and residential areas cover 46% of the watershed, located mainly in the upstream portion of the watershed. Forested areas, which make up 28% of the watershed, are located mainly in the downstream reaches. The remaining 26% is agricultural land. Stroubles Creek is a tributary of the New River, which drains portions of North Carolina, Tennessee, Virginia, and West Virginia before discharging to the Ohio River. The Stroubles Creek watershed lies in the Valley and Ridge physiographic province and is characterized by karst topography. Stroubles Creek was originally listed on the 303(d) TMDL priority list in 1996 for an aquatic life use impairment based on a benthic macroinvertebrate assessment. Sediment was identified as the primary stressor.
Location of Stroubles Creek watershed
The experimental reach used in this study is located on the Virginia Tech Foundation’s Heth Farm within Blacksburg town limits, approximately 2.5 km downstream from the Virginia Tech Duck Pond, located on the main campus. The drainage area used in this case study includes 1,440 ha, almost all of which lie in the Town of Blacksburg. The reach was selected because it is experiencing severe bank retreat. The upstream terminus of the study reach is located just downstream of a small tributary and the modeled section extends 500 m downstream. To simplify modeling procedures, the study reach was chosen to avoid hydraulic structures and tributary inflows.
A natural levee separates the incised channel from floodplain areas. The immediate area surrounding the experimental reach is pasture grazed by non-lactating dairy cattle, but the overall contributing area is dominated by residential and urban development. Cattle have unlimited access to the stream corridor and contribute to streambank instability. The floodplain vegetation surrounding Stroubles Creek on Heth Farm is composed of cool season grasses. Pocket wetlands formed in abandoned oxbows harbor zones of wetland vegetation, such as rushes (Scirpus sp.). Abandoned meander bends are visible on high resolution aerial photography suggesting past channel straightening.
The Stroubles Creek watershed has been influenced by human activity since the 18th century. Upland sediments, exposed by logging and agricultural activities, were eroded and transported downslope to floodplain areas. Significant silt deposits are now visible in Stroubles Creek bank profiles, with deposit depths ranging from several centimeters to over 50 cm. These fine textured legacy sediments are particularly susceptible to weakening and erosion due to subaerial processes. A clearly defined layer of McGary silt loam is visible on exposed bank faces throughout the reach. Ongoing development and past disturbance have resulted in an experimental reach characterized by unstable actively eroding stream banks. Downcutting has disconnected Stroubles Creek from the floodplain creating banks as high as 2.0 m and resulting in increased flow velocities and boundary shear stresses. Portions of the channel have downcut to an underlying erosion resistant saprolite layer. Underlying bedrock now prevents further downcutting and limits bed sediment supply. This natural bed control forces the stream to adjust laterally as it attempts to convey larger peak discharges associated with urban development.
Urban development has significantly altered the natural hydrology of Stroubles Creek. Changes in discharge occur rapidly following periods of rainfall. Observations indicate a two-hour time of concentration. Urbanization has also intensified differences between baseflow discharge and peak discharges. Baseflow discharge ranges from 0.03-0.17 cms, while bankfull discharges are approximately 8.5-9.9 cms.
Objectives & Methodology
The objective of this study was to compare predicted sediment loadings from in-stream sources as predicted by GWLF, SWAT, and CONCEPTS. Model comparisons were also contrasted with channel degradation estimates derived from a system of erosion pins and scour chains.
GWLF Model:
This case study utilized ArcView GWLF version 2005b software for channel degradation simulation. GWLF simulation was broken into two phases: a channel erosion sensitivity analysis phase and a one-year channel erosion prediction phase, concurrent with monthly erosion pin measurements. The sensitivity analysis of GWLF to flow and channel parameters was performed before the GWLF case study simulation.
SWAT Model:
The July 2005 update of SWAT 2000 (B. Narasimhan, personal communication, August 3, 2006), was used in the model comparison case study. The July update of SWAT 2000 includes corrections to the sediment routing algorithms. The ArcView SWAT 2000 (AVSWAT 2000) GUI was used to parameterize the model and create the input files. Watershed data input requirements for SWAT are broken down into three groups based on watershed subdivision structure: watershed, subbasin, or Hydrologic Response Unit (HRU) level parameters (Neitsch et al., 2002). The SWAT parameter set used in this research was adapted from the Wagner (2004) SWAT simulation. Necessary modifications to the Wagner (2004) parameter set, such as subbasin redistribution, were made using data available in the Stroubles Creek TMDL (Mostaghimi et al., 2003). Before running the 2005-2006 SWAT simulation a sensitivity analysis was performed.
CONCEPTS Model:
The CONCEPTS reach-scale model was used to simulate channel degradation for the Stroubles Creek Heth Farm study reach and compare the output with GWLF and SWAT. The objective of this effort was to evaluate if a more process-based, reach-scale model, coupled with a watershed-scale model, more accurately predicts channel degradation. A sensitivity analysis of CONCEPTS channel parameters was performed before simulating the case study scenario.
Monitoring Data:
In conjunction with the stream degradation modeling effort, an erosion pin monitoring study was conducted to quantify streambank retreat within the experimental reach. Streambank retreat rates were measured using erosion pins and scour chains on the 500-m long reach. Pins made from #316, 6-mm diameter stainless steel rod stock measuring 50 cm in length were marked on one end with 2.5 cm of brightly colored heat-shrink tubing and placed on actively eroding stream banks in a systematic 10-m horizontal and 0.3-m vertical grid. In total, 268 erosion pins and 7 scour chains were initially installed along the 500-m reach.
The point data collected through erosion pin monitoring was transformed into monthly mass sediment loadings to the stream. Point data were extrapolated to create an average predicted retreat surface. In an effort to limit extrapolation error, pin data were only applied to the continuous bank surface surrounding the pin, i.e. the directly connected, actively eroding portion adjacent to the pins. A measuring tape was used to estimate bank segment height and length. The erosion pin measurements were compiled and averaged for each individual bank. The average segment-specific retreat rate was multiplied by an estimated segment surface area. The soil volume was then multiplied by an average bulk density of 1300 kg/m3 as reported by Wynn, 2004 and Henderson, 2006.
Results and Conclusions
Sediment loading to the stream from bank retreat was estimated as 41 tonnes/yr, based on erosion pin measurements. GWLF, SWAT, and CONCEPTS predicted stream channel sediment contributions of 8 tonnes/yr, 1500 tonnes/yr and 4 tonnes/yr, respectively. Theil-Sen non-parametric simple linear regression was used to test agreement between monthly model predictions and erosion pin estimates. No significant agreement was found between any model predictions and measured retreat, using a conservative α-value of 0.2. GWLF model predictions underpredicted measured channel degradation, but most closely approximated observed data. This result is likely due to similarities in climate and watershed characteristics for the Stroubles Creek watershed and the Pennsylvania watershed used in the empirical model development. SWAT predicted retreat rates exceeded measured values by two orders of magnitude. This result is explained by the inability of SWAT to predict daily flow and sediment discharge. Highly sensitive channel degradation parameters and the lack of calibration data also contributed to SWAT simulation error. CONCEPTS simulation predicted monthly retreat rates slightly less than GWLF. The lack of agreement between CONCEPTS simulation and observed data was mainly the result of limited input data availability. SWAT daily discharge predictions were used as CONCEPTS input data and likely contributed to poor model agreement. Poor estimation of sensitive sediment input parameters may have also contributed to underpredictions by CONCEPTS. Results showed the potential of screening-level watershed models in channel degradation prediction and the importance of flow and sediment time series discharge data in detailed process-based simulation. The limited flexibility of the GWLF channel degradation algorithm makes it unsuitable for evaluating the effects of stream restoration. SWAT and CONCEPTS should only be used for evaluation if appropriate input data are available.
References
Henderson, M. B. 2006. Changes in streambank erodibility and critical shear stress due to surface subaerial processes. MS thesis. Blacksburg, VA: Virginia Tech. Department of Biological Systems Engineering.
Mostaghimi, S., Brian Benham, Kevin Brannan, Theo A. Dillaha, III, Rachel Wagner, Jeff Wynn, Gene Yagow, Rebecca Zeckoski. 2003. Benthic TMDL for Stroubles Creek in Montgomery County, Virginia. Prepared for VADEQ, Virginia Department of Environmental Quality. Richmond, Virginia.
Neitsch, S. L., J. G. Arnold, J. R. Kiniry, R. Srinivasan, J. R Williams. 2002. Soil and Water Assessment Tool, User’s Manual. Blackland Research Center, USDA-ARS, Temple, Texas.
Wynn, T. M. 2004. The effects of vegetation on stream bank erosion. PhD dissertation. Blacksburg, VA: Virginia Tech. Department of Biological Systems Engineering.
Additional Project Information
More detailed information on this case study can be found in the project’s final report available here.
Glossary
A number of terms are used to describe the loss of sediment from channel boundaries. These terms include channel erosion, streambank retreat, streambank erosion, scour, incision, and downcutting. For consistency and clarification, the following list defines the terms used on this web site and in the project report. Several of these definitions are based on a 1997 paper by Lawler et al. (Lawler, D. M., C. R. Thorne, and J. M. Hooke. 1997. Bank erosion and instability., Thorne, C. R. , R. D. Hey, and M. D. Newson, eds. Applied Fluvial Geomorphology for River Engineering and Management. John Wiley & Sons: Chichester. 138-172.)
| Bedload | Larger particles in bed material that are transported either by rolling, sliding, or saltation. |
| Channel degradation |
|
| Deposition |
|
| Equation | Matematcial relationships which provide the theoretical framework for model processes |
| Formula | see equation |
| Mass wasting |
|
| Model |
|
| Program | see model |
| Relationships |
see equation |
| Scour |
Direct removal of soil particles or aggregates from the streambed by stream flow |
| Sediment transport capacity |
Maximum amount of sediment a stream can carry at a give discharge |
| Software | see model |
| Streambank erosion |
Removal of soil particles or aggregates from the streambans due to stream flow |
| Streambank retreat |
Net recession of a streambank due to the combination of subaerial processes, fluvial erosion, and mass wasting |
| Subaerial erosion |
Removal of soil particles or aggregates due to soil weakening by subaerial processes |
| Washload |
Particles that remain in suspension |
© VT-BSE-TMDL Center 2006 | Updated 28 August 2006
Funded by US EPA Grant AW-83238501 Top
