Skip to main content
ARS Home » Pacific West Area » Boise, Idaho » Watershed Management Research » Research » Publications at this Location » Publication #330916

Research Project: Assessment, Conservation and Management of Rangelands in Transition

Location: Watershed Management Research

Title: Parameterization of erodibility in the Rangeland Hydrology and Erosion Model

Author
item AL-HAMDAN, OSAMA - Texas A&M University
item Pierson, Fred
item Nearing, Mark
item Williams, Christopher - Jason
item Hernandez, Mariano
item NOUWAKPO, SAYJRO - University Of Nevada
item Weltz, Mark
item SPAETH, KENNETH - Natural Resources Conservation Service (NRCS, USDA)

Submitted to: International Rangeland Congress
Publication Type: Proceedings
Publication Acceptance Date: 12/28/2015
Publication Date: 7/15/2016
Citation: Al-Hamdan, O.Z., Pierson Jr, F.B., Nearing, M.A., Williams, C.J., Hernandez Narvaez, M.N., Nouwakpo, S.K., Weltz, M.A., Spaeth, K.E. 2016. Parameterization of erodibility in the Rangeland Hydrology and Erosion Model. In: Proceedings of the X International Rangeland Congress, July 18-22, 2016, Saskatoon, Saskatchewan, Canada.

Interpretive Summary:

Technical Abstract: The magnitude of erosion from a hillslope is governed by the availability of sediment and connectivity of runoff and erosion processes. For undisturbed rangelands, sediment is primarily detached and transported by rainsplash and sheetflow (splash-sheet) processes in isolated bare batches, but sediment generally only travels a short distance before deposition. On disturbed rangelands, bare ground is commonly more extensive and runoff and erosion rates are higher. Increased erosion following disturbance occurs due to a shift from splash-sheet to concentrated-flow-dominated processes. Amplified runoff following disturbance transports splash-detached sediment further downslope into concentrated flow paths with high flow velocity, sediment detachment, and transport capacity. On long-disturbed sites, years of soil loss can limit sediment availability and soil erosion. In contrast, recently burned landscapes typically have ample sediment available and generate high erosion rates. This paper presents recent advancements in hillslope erosion prediction by the Rangeland Hydrology and Erosion Model (RHEM) that accommodate a wide range of vegetation, ground cover, and soil conditions and the associated dynamics in runoff and erosion. The RHEM tool is a process-based model that was developed specifically for predicting hillslope runoff and erosion on rangeland ecosystems. The advancements presented here required development of empirical equations to predict a splash-sheet erodibility parameter (Kss) for undisturbed and disturbed conditions. The data used for developing and evaluating the erodibility parameter equations were obtained from rainfall simulation databases maintained by the USDA-Agricultural Research Service. The data span undisturbed and disturbed conditions. Multiple stepwise linear regression analysis was used to derive the relationship between erodibility as dependent variable and ground and canopy cover attributes, slope, and soil texture as independent variables. Piecewise (segmented) regression analysis was applied where two continuous relationships between the log-transformed erodibility and the independent variables were fitted to improve the linear relationship. Applicability of the erodibility parameterization approach in RHEM was evaluated using a test of percent bias (PBIAS). A single generalized and four cover-type (bunch grass, sod grass, shrub, and forb) equations were developed for predicting splash-sheet erodibility. Applying piecewise regression, the best two-piece regression (R2 = 0.708) was obtained with a ground cover of 0.475 as the break point. The break point is consistent with that commonly identified as the 50-60% bare ground point in which small changes in bare ground promote large increases in sediment yield. Dividing the data into four groups based on the dominant vegetation community minimally improved the coefficient of determination (R2=0.713). The single versus vegetation specific approaches did not change the performance of prediction significantly. The overall performance of the Kss approach in RHEM was satisfactory with a PBIAS of 41.8, and the model predictions were able to match more than 50% of measured sediment yield even for disturbed sites. Model performance was slightly improved through a multiple parameterization approach using the Kss parameterization described in this study and the concentrated flow parameterization suggested by Al-Hamdan et al. (2015). The dual parameterization approach is recommended for conditions with abrupt disturbance (e.g., fire) on steep slopes with ample loose sediment. Our results demonstrate a single erodibility approach in RHEM is effective in predicting erosion across a wide range of conditions. A dual parameter approach, splash-sheet and concentrated flow, is only needed for cases of abrupt disturbance with steep slopes and ample sediment (e.g., imme