Stream Bank Erosion: In Situ Flume Tests K. Debnath 1 ; V. Nikora 2 ; and A. Elliott 3 Abstract: A new technique for testing the erodibility of cohesive stream banks using an in situ flume is presented. The erosion rate is estimated from direct measurements of bed surface elevations by acoustic sensors. The sediment resuspension rate is obtained using data on sediment concentrations measured by optical backscatter sensors and from water samples. The bed-load contribution to the total erosion rate is evaluated from the conservation equation for sediments. Temporal patterns of erosion and resuspension rates are studied employing stepwise increments of bed-shear stress. The data show that bed load plays a significant role in cohesive bank erosion. The data analysis suggests that erosion and resuspension thresholds observed in experiments were very low or equal to zero. The data support the power type equation for the erosion and resuspension rates with bed-shear stress as the key factor. The data also highlights the potential importance of mud content and water content on erosion. DOI: 10.1061/ASCE0733-94372007133:3256 CE Database subject headings: Cohesive sediments; Alluvial streams; Bank erosion; Bedload; Field tests; Flumes; In situ tests. Introduction Quantification and prediction of bank erosion requires knowledge of the dependence of the erosion rate on flow and bank material characteristics. Alluvial bank materials containing silt and clay sized particles are in general cohesive and resist entrainment pri- marily through interparticle electrochemical bonding rather than through the immersed weight of particles ASCE Task Committee 1998. These materials often have a continuous spectrum of par- ticle sizes and aggregation, ranging from single grains to large flocs or aggregates containing millions of grains. The aggregation may be enhanced by an increase in the number of clay-sized particles, which are primarily responsible for cohesive properties, as they provide more contact points for network formation McAnally and Mehta 2000. Therefore, erosion of cohesive bank materials often occurs through entrainment of aggregates rather than single particles. As a consequence, the quantification of the entrainment processes on the surface of these cohesive bank ma- terials is complex. The complexity is further compounded by the presence of microbial-organisms, fauna, macroflora, decaying or- ganic matter, organic secretions and biogenic structures, and their spatial and temporal variability. It is well known that cohesive materials change their proper- ties considerably during transportation from field to laboratory e.g., Black and Paterson 1997, and therefore, testing of erodibil- ity of bank materials in situ is preferable. Many studies have addressed fundamental aspects of erosion using in situ devices with particular focus on cohesive bed sediments e.g., Amos et al. 1992; Maa et al. 1993; Houwing and van Rijn 1998; Ravens and Gschwend 1999; Aberle et al. 2004. Erosion of cohesive banks is different from bed erosion as the processes of bank materials formation are different. In fact, the bed materials in most cases have shorter time scales for erosion-deposition-consolidation cycles compared to those for bank materials, which depend on history of flood events. Only a few in situ studies using jet de- vices e.g., Rheaume et al. 2002 have been conducted in relation to bank materials. The jet devices have a disadvantage that the flow field they generate does not represent the actual flow condi- tions associated with bank erosion. The advent of straight benthic in situ flumes e.g., Ravens and Gschwend 1999; Aberle et al. 2003; Grace et al. 2003 made it possible to study the erosion characteristics of natural beds under nearly natural conditions. However, no application of in situ flumes for studying the erod- ibility of stream banks has been reported yet. There are two rea- sons for this: 1 it is very difficult to place in situ flumes on steep stream banks for tests; and 2 known in situ flumes normally operate under submerged conditions, while their use on dry stream banks is problematic. Thus, to characterize the erodibility of unsubmerged banks with in situ flumes, a special technique had to be developed and tested. Such a technique should also account for potential effects of resuspension and bed load in the overall erosion process. In this paper we describe a new quasi in situ technique, and apply it to study erodibility of stream banks at five representative stream sites of the Auckland region in New Zealand. The sites were selected to cover a range of catchment sizes, geomorphic settings, geographic locations, and basement lithology. At each of the five sites, two erosion tests were conducted. Measurement Instruments The NIWA-II in situ flume developed by the National Institute of Water and Atmospheric Research NIWA used in our study is 1 Assistant Professor, Dept. of Applied Mechanics, Bengal Engineering and Science Univ., Shibpur, Howrah 711103, India; formerly, Postdoctoral Research Fellow, National Institute of Water and Atmospheric Research NIWA, P.O. Box 8602, Christchurch, New Zealand. 2 Professor, Dept. of Engineering, Univ. of Aberdeen, Aberdeen, AB24 3UE, U.K. E-mail: v.nikora@abdn.ac.uk 3 Manager of Ecosystem Modeling Group, National Institute of Water and Atmospheric Research NIWA, Hamilton, New Zealand. Note. Discussion open until November 1, 2007. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be filed with the ASCE Managing Editor. The manuscript for this paper was submitted for review and pos- sible publication on December 23, 2005; approved on December 29, 2006. This paper is part of the Journal of Irrigation and Drainage En- gineering, Vol. 133, No. 3, June 1, 2007. ©ASCE, ISSN 0733-9437/ 2007/3-256–264/$25.00. 256 / JOURNAL OF IRRIGATION AND DRAINAGE ENGINEERING © ASCE / MAY/JUNE 2007