Submitted to: Laboratory Publication
Publication Type: Government Publication
Publication Acceptance Date: January 21, 2010
Publication Date: N/A
Interpretive Summary: The South Branch of the Buffalo River is located within the Red River Basin, MN. In 1996 the Minnesota Pollution Control Agency (MPCA) did a study that showed that in some of the rivers in this basin, high levels of sediment and bacteria were polluting the rivers. Observations of the length of the South Branch of the Buffalo River included in this study (from the point at which Whisky Creek joins the South Branch of the Buffalo, to 31 km upstream), suggested that the banks are unstable in many locations, and much of the sediment that is polluting these rivers may be coming from the streambanks. The aim of this report is to calculate how much sediment is coming from the streambanks of the South Branch of the Buffalo River, and what percentage of the overall sediment in the channels comes from the banks versus other sources such as agricultural fields. Fiver sites were selected along the part of the river being studied. At each of these five sites measurements were taken to calculate soil strength, and soil erodibility so that these values could be used in a computer model (BSTEM) that gives the stability of the streambanks at different times and flow depths. Results of the computer model (BSTEM) showed that over a number of years ranging from very wet to very dry, volumes of sediment eroded from the streambanks at each site ranged from 42 to 1003 m3. The bank materials at each site had high cohesion values, and can resist bank failures unless the base of the bank is eroded by the flow of the river to a point where undercutting occurs. The contribution of sediment from streambank erosion for an average flow year was estimated to be 26.7 % of the suspended sediment measured at the closest USGS monitoring gage. In a very wet year the contribution from banks was estimated to be 43.2 % of the total suspended-sediment load measured at the gage. The contribution from the banks was higher in this study than in a previous study by Lauer et al. (2006). The reason for this is that the Lauer et al. (2006) study looked at a relatively dry period, but this study also included the last fifteen years, which have been much wetter. As bank erosion is higher in wetter years, this explains why the amount of sediment coming from the streambanks is higher in this study than in the previous study. In drier years, the values of sediment coming from the streambanks may be lower and may be closer to the values from the previous study of Lauer et al. (2006).
Technical Abstract: The South Branch of the Buffalo River is part of the larger Red River Basin, MN. In 1996 the Minnesota Pollution Control Agency (MPCA) performed water quality assessments for selected rivers and lakes in the Red River Basin, with impairment of streams primarily being found to be caused by high levels of sediment and bacteria, and low levels of dissolved oxygen. Observations along the study reach of the South Branch of the Buffalo River (extending from the confluence with Whisky Creek, to 31 km upstream) indicated that the river’s streambanks are unstable in some reaches and are a potential source of a significant proportion of the sediment causing the suspended sediment issue in this part of the Red River Basin. The main objective of this study was to determine loadings of sediment from streambank erosion along the study reach of the South Branch of the Buffalo River, MN, and to estimate the percent contribution to overall suspended sediment yields from streambank erosion versus upland sources. Five study sites were selected from the 31 km study reach, to act as representative conditions for the entire reach. At each site, data pertaining to geotechnical strength and hydraulic resistance were measured to use as input data to the Bank Stability and Toe Erosion Model (BSTEM version 5.1). Results of the BSTEM analysis for a range of percentile flow years (99th, 95th, 75th, 50th, and 25th) showed that predicted eroded volumes of sediment emanating from streambanks generally decreased non-linearly from the 99th percentile flow year to the 25th percentile flow year, although some exceptions did occur. Predicted volumes of sediment eroded from the streambanks at each site ranged from 398 to 740 m3 of sediment per 100 m reach during the 99th percentile year, and from 42 to 1003 m3 for all of the years modeled. The bank materials tested at the five sites had relatively high cohesion values, but were susceptible to erosion by hydraulic scour. Without scour of the toe and undercutting of the banks, the high cohesion of these streambanks makes mass failures unlikely. In the wetter years modeled, cantilever failures resulted when the profile became so undercut that the bank’s driving forces exceeded the resisting forces, even with the cohesion of the silt clay loams and root-reinforcement provided by grass species. The contribution of sediment from streambank erosion for an average annual flow year was estimated to be 26.7 % of the measured suspended sediment at the closest gage (05061500). For the 99th percentile flow year the contribution from banks was estimated to be 43.2 % of the total suspended-sediment load at the gage. The values estimated in this study for percent contribution to suspended sediment loads from streambank erosion (26.7 to 43.2 %), were considerably higher than the 11% predicted in the Lauer et al. (2006) study of the Red River Basin. However, when the rainfall and water yield records for Ecoregion 48 were analyzed (Klimetz and Simon, 2009), it could be seen that the period of record used in the Lauer et al. (2006) study (1939-1991) was a drier period, with the last 15 years that were also included in this study (1945 – present), being part of a particularly wet period for this ecoregion (Klimetz and Simon, 2009). It is perhaps unsurprising then, that the estimates of contributions from bank erosion in this study, are higher than the 11% given in the Lauer et al. (2006) study as this study encompassed a wetter period of record, and more bank failures occur during wetter years. In drier years, therefore, the contribution from streambanks may be lower, and may be closer in value to the estimate of 11% provided by Lauer et al. (2006).