Water Resources Research Act Program

Details for Project ID 2015OK318B

Quantifying Streambank Erosion and Phosphorus Load for Watershed Assessment and Planning

Institute: Oklahoma
Year Established: 2015 Start Date: 2015-03-01 End Date: 2016-02-29
Total Federal Funds: $25,000 Total Non-Federal Funds: $50,000

Principal Investigators: Daniel Storm

Project Summary: The Illinois River watershed in eastern Oklahoma is a designated scenic river and a highly valued system visited by many Oklahomans every year. The Illinois River along with the Baron Fork Creek is on the 303(d) list of impaired waters due to elevated phosphorus (P), which is caused in part from historic poultry litter application and wastewater treatment discharges. Since over 80% of the litter produced within the watershed is now exported outside the watershed and wastewater treatment plant discharges have improved dramatically, they are no longer the largest P sources in the watershed. Current Soil and Water Assessment Tool (SWAT) modeling has shown that pasture and elevated soil test P are now the largest P sources. Recent research on the Baron Fork Creek, a major tributary of the Illinois River and also a designated scenic river, has shown that streambank erosion is also a significant P source in the watershed (Miller et al., 2014). Past SWAT modeling efforts of the Illinois River watershed have ignored the contribution of stream banks as a P source due to lack of field data and model limitations. With recent modifications to the SWAT model and data collected by Miller et al. (2014), it is now possible to estimate P contributions from stream banks in the Baron Fork Creek watershed. This will not only improve our understanding of the Baron Fork Creek watershed and other streams in the Ozarks, but also contribute and improve the well-known and widely used SWAT model. The first objective of this research is to estimate streambank erosion at the ten reaches from the Miller et al. (2014) study using the SWAT model, and compare the simulated eroded volume of sediment to the observed erosion. This will be accomplished by: 1) Setting up a SWAT model for the Baron Fork Creek watershed using the latest available data. 2) Calibrating and validating flow and P using the new SWAT in-stream P subroutine, ignoring streambank erosion. 3) Predicting stream bank erosion using SWAT default parameter values (bank height and width, critical shear stress, erodibility coefficient, and bank composition) and Yang’s gravel transport equation. 4) Estimating the volume of eroded streambank using ArcGIS 10.1 and aerial images from 2003-2013. 5) Calibrating the stream bank erosion by incorporating the measured parameters from Miller et al. (2014), using an improved excess shear stress equation, and by considering the sinuosity of the reaches. The second objective addresses the 2015 Water Research Funding Priority 3f by extending the streambank erosion estimates to the entire Baron Fork Creek watershed. After obtaining satisfactory results from the 10 reaches, we will extend the streambank parameters to each of the reaches on the Baron Fork based on the aerial images and their current riparian protection. This SWAT simulation will provide annual and event based erosion estimates on the Baron Fork. After modifying the SWAT code, the quantity of P from the eroded stream banks will be added to the benthic P in the in-stream P subroutine and the model will be re-calibrated for P. We expect the inclusion of P from stream banks to improve the simulation results and therefore give us a defendable estimate of sediment and P added to the Baron Fork Creek each year from streambank erosion. Methods used in this research can be applied to other stream systems in eastern Oklahoma, such as the Illinois River and Spavinaw Creek. This will give policy makers a science based estimate of the quantity of sediment and P from stream bank erosion, which will allow them to make informed decisions on selecting restoration activities. To date, gravel stream banks with extensive erosion, such as that observed in the Baron Fork watershed, has not been simulated with the SWAT model. Therefore, this research will provide invaluable input on the current state of the streambank erosion processes currently available in SWAT and help identify the areas that need improvement. A summary of the benefits from the proposed research include: 1) Provide local, state and federal agencies with accurate estimates of streambank erosion and P contributions for the Barren Fork Creek watershed, which will allow them to make informed decisions on funding priorities and selecting the most cost effective upland and/or streambank restoration activities, 2) Improve TMDLs and watershed based plans by incorporating defendable estimates of streambank derived sediment and P, 3) Test and assess a process-based model that will be incorporated into SWAT and will be applicable to other watersheds throughout the world, and 4 ) Provide recommendations to watershed modelers and managers to focus data collection and parameter estimation efforts on the most critical streambank erosion parameters; thus providing more accurate model predictions.