Current regulations in North Carolina require sediment basins to be designed based on watershed size, regardless of watershed characteristics. For each disturbed acre, the basin needs to have 1,800 cubic feet of sediment storage and a surface area of 325 square feet if a surface outlet is used. Another approach to designing a basin would be to determine the expected runoff and sediment loads based on watershed characteristics, including soil type and slopes. By using local soil and weather data, the basin size would be adjusted to account for expected sediment loads. To determine if this is a valid approach, we will collect soil and topography data in at two basin watersheds on DOT projects and monitor sediment and flow over the landscape and into and out of the basins. This will be done during four critical phases of the site: preconstruction, mass grading, final grade, and post-construction.
The information collected on the sites will then be used to determine the proper questions that really need to be addressed in using RUSLE to do conservation planning by estimating erosion and delivery on such sites. For example, in what phases of construction will erosion be worst, and will that occur on the hillslopes or in the channels? Do high sediment loads delivered to the channel tend to be deposited there, only to be re-eroded by later events when hillslope erosion has diminished? How do these patterns change over time? Only by answering such questions will it be possible to make optimum use of RUSLE in planning for these sites.
There will be five activities for each site: Survey the landscape at critical times (grading phase, storm events).Collect representative soil samples for particle size analysis and conduct infiltration tests. Install flow measurement and sampling equipment at the inlet to the basin (or watershed outlet for preconstruction. Monitor storm flow and sediment delivery to the outlet. On a storm basis, collect qualitative classification information (e.g., photos, field notes) describing the visible signs of erosion and sediment movement over the landscape. Quantify using laser surveys. Modifications/additions will be made to the RUSLE code as necessary to allow it to better represent the critical controlling situations. This will provide the background data needed to determine how RUSLE2 can be utilized to provide basin designs appropriate to local conditions, and will ultimately result in an improved RUSLE version specifically designed to better address site conditions.
The purpose of this project was to determine actual sediment yield at sediment basin inlets and to compare that to predicted yields estimated with RUSLE2. This involved monitoring four basins on a Piedmont project and one on a Coastal Plain project. Water samples were obtained during storm events at the inlet and outlet of each basin and analyzed for turbidity and total suspended solids (TSS). The flow data was used to calculate the amount of sediment reaching each basin. In addition, detailed surveys of the basin catchments were obtained using a LiDAR (Light Detection and Ranging) instrument. The amount of sediment delivered was highly related to storm intensity but not amount of rainfall. From observations of field operations, the presence of active grading also resulted in higher sediment yields. Peak flow was highly predictive of total sediment load, turbidity, and TSS. Total flow volume was also correlated with total sediment load and turbidity, but less with TSS. Peak flow was correlated with peak rain intensity but not with rainfall amount. Peak intensity was negatively correlated to basin performance in reducing TSS and turbidity. All of these results suggest that the current practice of using peak rain intensity for a design event to estimate runoff flow is appropriate. The relationship between actual sediment yield and RUSLE2 predicted sediment was very poor for 3 of the 5 basins studied due to erosion within the ditches. Tests of RUSLE2 sediment yield using either actual site surveys using LiDAR or slope factors from the clearing and grubbing (CG) or final grade (FG) were poorly correlated with measured sediment yield at the basin inlet. The CG slopes tended to overpredict and FG slopes underpredict actual sediment yields. The two basins where actual and RUSLE2 predicted sediment yield were similar had diversion ditches which were relatively stable and non-eroding. This suggests that stabilizing ditches and slopes quickly will greatly reduce sediment loading and may allow predictive models such as RUSLE2 to be used.
The current method of design for sediment basins on NCDOT highway construction sites only takes into account a few of the factors that can affect true sediment yield from a catchment. Using an empirical or process based model could be a faster and more accurate method of sizing these sediment basins in order to achieve a desired performance.
Based on the observations and results of this research, RUSLE2 would not be a good method for estimating sediment yield and sizing sediment basins on NCDOT highway construction sites. The changing dynamic of the landscape, the differences in catchment management, and the lack of ability that RUSLE2 has in estimating channel erosion are all reasons that the estimate from RUSLE2 can be an over- or under- prediction of the sediment yield on site. Because of the landscape dynamics and unpredictable rain patterns, there was no clear evidence that RUSLE2 would typically under- or over-predict the measured sediment yield.
Furthermore, RUSLE2 can only estimate a sediment load and therefore would only be able to estimate the sediment storage volume. Neither sediment transport factors nor concentrated flow erosion are included in the model. The model also does not determine deposition in the channels due to the effects that BMPs, such as check dams, and the effects on sediment yield at the basin inlet.
The similar sediment yield and RUSLE2 estimate on Basin 5.10 B and at Goldsboro emphasizes the importance of protecting the ditches leading into the basins, and that RUSLE2 can be more accurate when the topography is known.
The topography in clearing and grubbing plans tended to produce much higher sediment yields in RUSLE2 than that of the final grade. During the transition, the catchment tended to change from relatively steep slopes near the basin to shallow slopes near the basin, longer slopes, and smaller catchments. In construction scenarios similar to these, it may be more conservative to use the CG plan topography to estimate sediment yield with RUSLE2.
