GEOMORPHIC PROBLEMS WITH IN-STREAM STRUCTURES USING NATURAL CHANNEL DESIGN STRATEGY FOR STREAM RESTORATION PROJECTS IN NORTH CAROLINA

Federal, state, local, and private funds are increasingly being committed to stream restoration projects in the United States, to stabilize stream channels, reduce bank erosion, and improve aquatic habitat. The majority of these projects utilize a method known as Natural Channel Design. As used by many restorationists, National Channel Design relies on the classification of channels into Rosgen type stream reaches and assigns a prescriptive restorative treatment. Hallmarks of this strategy employ in-channel rock structures such as cross-vanes, J-hooks, and rock weirs as well as bank treatments such as rootwads, revetments, and live stakings. We are in the midst of a comprehensive 3-year survey of restoration projects in North Carolina completed within the past 7 years, most constructed using Natural Channel Design concepts. The data are intended to provide guidelines for design of future restoration projects. Preliminary results of field surveys show that along some rivers more than 70% of in-stream structures have sustained significant damage; structures in some cases totally destroyed by the first significant flood. Damages vary considerably, but generally fall into several categories: 1) flood erosion around the landward side of J-hooks and cross-vanes, accelerating bank erosion locally and stranding the rock vanes as ineffective mid-channel islands; 2) complete removal of rock vanes by floods; and 3) diversion of flood flows onto floodplains beside in-stream structures. Along one stream an extensive set of structures was buried by several feet of alluvium during channel aggradation that followed suburban development in the upstream watershed. These projects appear to perform reasonably well under low flow to bankfull conditions; the vast majority fail or suffer major damage during bankfull to flood stage. While the structures modify flow patterns as designed during low flow, they appear to simply act as roughness elements during high flow. Many times, they clearly accelerate bank erosion along the stream system. Moreover, most projects fail to consider changes in sediment and water yield that have occurred, or are occurring, due to historical land use changes and/or major flood events in upstream portions of the drainage basins. In these cases, streams may be in a protracted phase of morphologic adjustment, rendering restorative structures virtually useless.