UVM Theses and Dissertations
Format:
Print
Author:
Borg, Jaron L.
Dept./Program:
Civil and Environmental Engineering
Year:
2010
Degree:
MS
Abstract:
With over 11,000 kilometers of streams in Vermont and extensive agricultural land use sediment associated nutrient loading represents a major pollution source. To improve land and waterway management practices further understanding of sediment transport and streambank stability is desirable. Two focus areas are presented in this work; a longitudinal study conducted at two eroding stream reaches to observe streambank stability mechanics in tributaries of Lake Champlain, VT. Work was also done to evaluate the use of the radionuclide ¹⁰Be as a fingerprint for suspended sediments in post glacial and temperate regions.
Assessment of streambank stability was preformed through a series of crosssectional surveys ys during the 2007, 2008 and 2009 field seasons. In combination with the surveys, geotechnical properties of the streambanks were measured using insitu and laboratory methods. These included soil shear strengths, densities, gradations, erosion characteristics and tensile reinforcement from vegetation. One cross-section at each of the stream reaches was instrumented to capture the time of a streambank failure should it occur and varying hydraulic conditions. Hydraulic conditions at the two reaches were recorded, collecting ground and stream-water elevations. To capture a streambank failure event a data logging tilt switch array was used. A condensed set of geotechnical properties were then used in a slope stability computer program to determine the stability of each cross section for the range of hydraulic conditions occurring at each study reach.
Suspended sediment samples were collected from seven different sampling points located· in the Lake Champlain watershed. Sampling sites were selected to represent a range of watershed characteristics; forested, agricultural, upland, lowland and impaired watersheds. Five sampling points were located in the Winooski River watershed; the remaining two were selected in an adjacent watershed to minimize geographical influences. Samples were taken during high water events when a majority of sediment is transported. To evaluate seasonal trends in sediment sources three sample sets were collected to represent a spring melt event, a summer storm event, and a/all storm event. An additional set was retrieved to observe changes in sediment sources on a daily timescale. Data collected were analyzed with respect to landuse characteristics and history of each watershed.
Rapid drawdown condition did not induce streambank failures at the study sites, because the groundwater levels in the banks followed stream water levels very closely. The low stream water level condition paired with the loss of matric suction from a rapid wetting event yielded the lowest computed factors of safety, which compared reasonably well with the recorded streambank failure event. . It is anticipated that the effects of freezing and thawing of streambank soils and ice flows in the streams, typical 'of humid temperate climate of Vermont, could be critical in evaluating streambanks in Vermont. Measurements of ¹⁰Be concentrations in suspended sediment ranged from 0.3 to 18.7*10⁸atoms/gram. Statistical differences in concentrations could not be seen on daily or seasonal timescales. Differences in ¹⁰Be concentrations showed contributing watersheds have statistically different suspended sediments, which correlate to forested and agricultural land area in the contributing watershed.
Assessment of streambank stability was preformed through a series of crosssectional surveys ys during the 2007, 2008 and 2009 field seasons. In combination with the surveys, geotechnical properties of the streambanks were measured using insitu and laboratory methods. These included soil shear strengths, densities, gradations, erosion characteristics and tensile reinforcement from vegetation. One cross-section at each of the stream reaches was instrumented to capture the time of a streambank failure should it occur and varying hydraulic conditions. Hydraulic conditions at the two reaches were recorded, collecting ground and stream-water elevations. To capture a streambank failure event a data logging tilt switch array was used. A condensed set of geotechnical properties were then used in a slope stability computer program to determine the stability of each cross section for the range of hydraulic conditions occurring at each study reach.
Suspended sediment samples were collected from seven different sampling points located· in the Lake Champlain watershed. Sampling sites were selected to represent a range of watershed characteristics; forested, agricultural, upland, lowland and impaired watersheds. Five sampling points were located in the Winooski River watershed; the remaining two were selected in an adjacent watershed to minimize geographical influences. Samples were taken during high water events when a majority of sediment is transported. To evaluate seasonal trends in sediment sources three sample sets were collected to represent a spring melt event, a summer storm event, and a/all storm event. An additional set was retrieved to observe changes in sediment sources on a daily timescale. Data collected were analyzed with respect to landuse characteristics and history of each watershed.
Rapid drawdown condition did not induce streambank failures at the study sites, because the groundwater levels in the banks followed stream water levels very closely. The low stream water level condition paired with the loss of matric suction from a rapid wetting event yielded the lowest computed factors of safety, which compared reasonably well with the recorded streambank failure event. . It is anticipated that the effects of freezing and thawing of streambank soils and ice flows in the streams, typical 'of humid temperate climate of Vermont, could be critical in evaluating streambanks in Vermont. Measurements of ¹⁰Be concentrations in suspended sediment ranged from 0.3 to 18.7*10⁸atoms/gram. Statistical differences in concentrations could not be seen on daily or seasonal timescales. Differences in ¹⁰Be concentrations showed contributing watersheds have statistically different suspended sediments, which correlate to forested and agricultural land area in the contributing watershed.