Water Resources Research Act Program

Details for Project ID 2016VT79B

System-wide rapid quantification of streambank erosion

Institute: Vermont
Year Established: 2016 Start Date: 2016-03-01 End Date: 2017-02-28
Total Federal Funds: $39,958 Total Non-Federal Funds: $80,160

Principal Investigators: Mandar Dewoolkar, Jarlath ONeilDunne, Donna Rizzo, Jeff frolik

Project Summary: Streambank erosion is estimated to account fo r30-80% of sediment loading into waterways. In many cases, this sediment is carrying important pollutants, such as phosphorus. Langendoen et al. (2012) recently completed a study involving extensive fieldwork and BSTEM (Bank Stability and Toe Erosion Model) modeling for the State of Vermont to estimate sediment loadings from streambank erosion in main stem reaches of Missisquoi River. Using the flow records between 1979 and 2010, they predicted that 36% of the total suspended-sediment load entering Missisquoi Bay (31,600 t/yr) was derived from streambank erosion. These estimates were based on "one-time", yet labor and resource intensive, field work conducted at 27 sites that were extrapolated to 110 km of stream length and across 30 years in time. Although this study demonstrated that the estimates of streambank erosion can be obtained at the watershed level, this approach requires tremendous resources. Recent approaches such as aerial LiDAR have proved effective for watershed level assessment, but airborne LiDAR data collection is costly. Terrestrial LiDAR is more affordable if the equipment is available, but the equipment is bulky (especially for reaching remote locations) and data collection is time consuming and limited to relatively small areas. Therefore, cost-effective approaches to reliably quantify streambank erosion at the watershed level have remained an elusive goal. Recent developments in Unmanned Aircraft Systems (UAS) provide opportunities for rapidly and economically determining streambank erosion and deposition at variable scales (from site specific to watershed scale). The objectives of the proposed study are to: (1) develop decision support tools to effectively acquire and process continuous streambank profiles using an affordable UAS; (2) compare the results at select sites from terrestrial and airborne LiDAR-based surveys; (3) develop and validate a methodology to reliably quantify annual system-level streambank erosion and deposition rates; and (4) develop and incorporate related educational modules for the University of Vermont (UVM) coursework and conduct professional development workshops for Vermont state and government personnel, and disseminate the results through publications. This proposal requests funding for the second year of this ongoing research effort. Considerable progress has been made to date (~7 months) since March 2015 in the first year of this project. A total of 18 km of stream reaches within the Mad River watershed and 2 km along the Winooski River were flown using the SenseFly eBee UAS during the spring and summer. Six streambank sites were simultaneously scanned using terrestrial LiDAR. Preliminary data processing for all six sites has been completed and shows good agreement between the UAS and terrestrial LiDAR data in areas not obscured by vegetation. Additional work planned for Year 1 includes repeat flights and scans after leaf off in fall 2015. Comparisons between spring and fall data are expected to show areas of change caused by erosion during summer months. Year 2 effort will include repeat flights and scans again in spring and fall, with additional flights in response to storm events. Analysis in Year 2 will also include similar comparison to the aerial LiDAR data expected to be released in spring 2016, and will allow assessment of the ability of UAS to reliably quantify annual streambank erosion and deposition rates at a watershed level. This work should have substantial impact on the understanding of bank stability and sediment input to Vermont streams. In particular, we will be able to provide a field-validated methodology that will allow reliable quantification of the contribution of streambanks to sediment loadings in waterways, using Vermont as the case study. The developed methodology will be cost-effective for measuring rate and quantity of streambank erosion and transferrable to other regions in and outside of Vermont.