Identifying Channel Erosion Hot Spots

Geomorphic assessments help identify problem areas in small urbanized watersheds.

The sedimentation of small reservoirs in urbanizing watersheds is a problem that plagues many communities throughout North America. As watersheds become urbanized, sediment loads in stream channels can increase due to rainsplash erosion of soils undergoing land-use change, streambank erosion can increase because of increased runoff, and bed incision can increase in response to stream channelization. Resulting downstream sedimentation has the potential to create conflict between sediment-generating areas and depositional areas of a watershed.

A typical case is that where a small reservoir (less than 100 acres) is located downstream of an urban watershed. Those cognizant of the reservoir, upon learning the high cost of sediment removal, often seek to blame an upstream entity in hopes of raising the needed funds to help pay for removing the sediment from their reservoir. To resolve a reservoir sedimentation issue without conflict, the City of Griffin (GA) Stormwater Department is taking proactive steps to locate and mitigate the sources of sediment that are filling the Griffin Country Club Lake.

The lake, a 40-acre impoundment constructed in 1968 at the lower end of the 2,700-acre Shoal Creek watershed at Griffin, is undergoing siltation at a rate not acceptable to the country club owners. Sets of gabion check dams had recently been installed at the reservoir inlet to trap sediment. The design called for dredging the sediment trap approximately once every eight to twelve months; however, the constructed traps required dredging every six to eight weeks. The flow of sediment into the lake far exceeded the design calculations, and the dredging expense far exceeded the allotted budget. To resolve the sedimentation issue, the City of Griffin Stormwater Department needed to locate and mitigate the sources of the sediment.

The Shoal Creek watershed is roughly split into an urban upper half and an agricultural lower half, with the upper half being mixed commercial and residential land use. The majority of the lower half belongs to the University of Georgia Agriculture Experiment Station, and land uses there historically include pasture, row crop, fallow, and forest. The remainder of the watershed is nearly 100% built out. The last major construction included separate Lowe’s and Super Wal-Mart plazas since 1999. During the development of these plazas, stormwater detention basins were designed and constructed to control runoff peaks generated from the parking lots and rooftops.

Two potential sources of sediment were hypothesized: filter backwash from the City of Griffin Drinking Water Treatment Plant (DWTP) being scoured from storage on the channel bed, and channel erosion of Shoal Creek’s main stem and its numerous tributaries flowing through the University of Georgia Agriculture Experiment Station.

Brant Keller, director of public works and stormwater for Griffin, seeking a more complete assessment of potential sediment sources, asked Tetra Tech to design an assessment plan that could be quickly carried out and provide some resolution to the sources of sediment. Tetra Tech proposed sending an applied geomorphologist, experienced in assessing stream channels for potential sediment sources, to make specific observations by walking the stream channels and interpreting aerial photos of the watershed. The ultimate goals were to determine the geomorphic state of Shoal Creek, determine sources of sediment to the lake, and create a stream channel erosion hot-spot map.

The first step in the project was to create an organized fieldwork plan based on available watershed data. Road maps and stream maps were used to measure the total length of streams to be assessed and to find access points to the streams. The maps were provided by the City of Griffin in the form of geographic information system (GIS) datasets, which also were useful in creating maps for presenting the results. The total length of the stream channels was estimated at 10 miles. From previous work on similarly sized streams, an assessment rate of 2.5 miles per eight-hour day was estimated. This indicated that the 10 miles of streams could be completed in four days.

Included with the GIS datasets were aerial photographs of the watershed shot in 1999. These photos proved immensely useful because they provided the necessary information to resolve discrepancies between field observations and the GIS streams and lakes data layer. Several stream reaches were inaccessible because they ran through fenced private property; however, the photos provided enough information to conduct a partial stream channel assessment of these reaches. Additionally, the aerial photos indicated areas that have undergone land-use change since 1999. The photos clearly showed the extent of land disturbance at the Lowe’s and Super Wal-Mart construction sites as well as several smaller-scale land development projects.

The second step was to put a geomorphologist in the field to conduct the channel stability assessment. Aerial photos, land-use maps, topographic maps, and anecdotal evidence from landowners can set the background for conducting an overall sediment source assessment. But to learn the specific locations where erosion is taking place—and the relative magnitude of erosion at each location along a stream network—an experienced geomorphologist is needed to assess, measure, and photograph the stream channels.

The channel stability assessment was made according to the Channel Evolution Model (CEM) (Simon 1989). The CEM predicts channel stability based on observable characteristics of the channel, including form, bed material, and riparian vegetation. Using this method was a good fit with the Shoal Creek project because the qualitative assessments can be quickly conducted, CEM is suited to alluvial streams, and Tetra Tech staff have strong experience applying this method to streams throughout the southeastern United States. The CEM was designed to describe stream channels that have undergone a disturbance and predict how the channels will continue to respond (Figure 1). Typical disturbances include channelization, change in discharge, and change in sediment load. These types of disturbances can raise or lower the amount of energy available in flowing water (stream power), which results in differences in how the streambed, banks, and transported sediment respond to that energy change. To accommodate changes in stream power, channels typically need to adjust the gradients along their beds. The gradient can be directly adjusted through bed incision or bed aggradation, or indirectly adjusted through changing the channel length through increased or decreased meandering rates. The CEM is a screening tool used to classify what stage a channel is in (stable, incising, widening, aggrading, or restabilized) and to predict if a channel is likely to become unstable, remain unstable, or become stable. Knowing the present and potential stability state of a stream channel is important for targeting areas susceptible to ongoing and future erosion.

Data collected during the CEM assessment include streambank heights, bed material, riparian vegetation, and present erosion processes. Geomorphically, each of these parameters indicates something about the present stability of the channel, as well as potential future stability trends, in the following ways:

Streambank height. Banks composed of unconsolidated materials such as clay, silt, and sand have a stable height limit based on the materials’ saturated strength. By noting bank heights and the presence (or absence) of bank mass wasting at locations throughout the watershed, one can estimate the stable bank height limit.

Bed material. The type and depth of bed material indicates if the streambed is downcutting, aggrading, or stable. If a bed is scoured to a hard substrate such as bedrock, hard clay, or cobbles, this indicates that there is either a lack of sediment or an excess of flow. If a channel is bedded with a deep layer of soft sand, this indicates either an excess of sediment or a lack of discharge to transport the sediment.

Riparian vegetation. The type, density, and age of riparian vegetation indicate channel stability in several ways. Banks without deep root penetration are more susceptible to mass wasting than are densely and deeply rooted banks. Perennial vegetation—typically trees and woody-stemmed shrubs—serve as indicators of stability where they are found. Woody vegetation has difficulty becoming established below the level of the one-and-a-half-year recurrence interval flood (effective channel forming discharge) and on not-yet-stable bars.

Bank erosion. Streambanks can erode by several different processes. They can scour away inch by inch, be undercut and then slough off, or mass waste in a mini-landslide. These different erosion processes indicate different rates of sediment supplied to the stream and also indicate different causes such as devegetation, channelization, or a change in discharge.

In addition to field notes and measurements of channel form and materials, photographs are taken and assessment locations are recorded for each site.

Photographs. Photos are taken as a record of channel form, bed material, bank material, vegetation, typical features, and unique features, and as a historical document of the stream. The photos are also used for comparing reaches relative to each other when determining which reaches exceed the threshold for being an erosion hot spot.

Location. Accurately recording the location of each assessment point is critical for communicating the results to the client. Global positioning system (GPS) technology provides the fastest and most accurate way of collecting location data that can be quickly mapped. Additionally, notes are taken about features in the area, such as location relative to bridges, numbered manhole covers, and tributary junctions.

The most effective way for a geomorphologist to collect these types of data is to walk on the bed of the channel. This provides the ability to view both banks in their entirety. The assessor looks for changes in the assessment parameters such as a transition from riparian buffer of dense woods to one that is clear cut, or from a streambed of hard clay to one that is sand. At each transition point an assessment is made of all parameters. The reasons for walking the streambed instead of the banktop are that often only the opposite bank can be clearly viewed from the banktop, and walking can be difficult through densely vegetated riparian zones, especially when briars and poison ivy dominate.

Luckily, most of the reaches of Shoal Creek and its tributaries were accessible. Several of the upper reaches of the tributaries were inaccessible because of the dense vegetative canopy over the channel. However, dense vegetation is a general indicator of channel stability, and small channels have less surface area exposed to erode; therefore it is probable that sediment supplied from these reaches is negligible compared to the quantities from the main channel of Shoal Creek and lower reaches of the tributaries. Access to the urban reaches was limited because of fences and private property. However, the large number of road crossings and lack of riparian vegetation made it easy to see long distances from the roads and therefore quickly complete assessments without needing to trespass.

During four days in the field, 112 sites were assessed along the main channel of Shoal Creek and 18 tributaries. Data from these sites were compiled into a table sorted in order from upstream to downstream for each tributary. Each stream reach was then classed as either erosion hot spot (Figure 2), erosion sensitive (Figure 3), or stable (Figure 4), based on the size of the area eroding and the erosion process. The various reaches were then plotted to create a hot-spot map based on the GPS data (Figure 5). The map, in conjunction with the data table and photos, creates a unified database showing where present and potential erosion concerns are within the watershed and the relative significance of each.

Several key characteristics of the watershed were learned from the assessment

• The lowest mile of Shoal Creek is aggrading, and much of the sediment being eroded from the upstream reaches is being deposited on the streambed and floodplains.
• The middle reaches crossing the University of Georgia property are stable because of the densely rooted vegetation protecting the riparian zone.
• The most severe stream channel erosion is occurring along the upper reaches within city limits. The cause appears to be bank destablization from lack of vegetation. Increased runoff peaks may also be a significant factor.
• A headcut on a tributary is threatening the stability of a 2-acre impoundment.
• Incision of the bed is not a problem in general because of frequent natural bedrock grade controls.
• No fine sediment deposits such as those historically discharged by the DWTP were found.
• Channels draining the Super Wal-Mart plaza are presently stable. The North Griffin Regional Detention Pond system, constructed in conjunction with the Super Wal-Mart, appears to be effectively handling increased runoff peaks from the increased impervious area.
• The only bare soil noted in the watershed includes both residential yards kept grass-free and urban construction sites less than 2 acres in size.

Advertisement

Conclusions
Although many of the characteristics of the watershed were already known by local landowners, they were not known collectively to the Stormwater Department. This study compiled these characteristics and thus provided a format for comparing the significance of each related to sediment supply in the watershed. Additionally, during the study, other avenues related to sediment generation were discovered for future exploration.

Ultimately, for the City of Griffin, the Griffin Country Club, and the University of Georgia, this study provides the needed background to begin targeting sediment sources and resolving the lake sedimentation issue.

Author's Bio: Nick Jokay is an applied geomorphologist with Tetra Tech in Atlanta,

Author's Bio: Brian Watson, P.E., is a project manager and water resources engineer with Tetra Tech in Atlanta,

Author's Bio: Brant D. Keller, Ph.D., is director of Public Works & Utilities for the City of Griffin, GA. He was instrumental in setting up the first stormwater utility in the state of Georgia. He is a member of Stormwater magazine’s editorial advisory board.