INTEGRATED ASSESSMENT AND ANALYSIS OF PHYSICAL LANDSCAPE PROCESSES THAT IMPACT THE QUALITY AND MANAGEMENT OF AGRICULTURAL WATERSHEDS
Location: Watershed Physical Processes Research Unit
Title: Scaling a representative storm sequence to estimate ephemeral gully erosion with RUSLE2
Submitted to: Federal Interagency Sedimentation Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: May 1, 2010
Publication Date: June 27, 2010
Citation: Dabney, S.M., Yoder, D.C., Vieira, D.A., Bingner, R.L., Wells, R.R. 2010. Scaling a representative storm sequence to estimate ephemeral gully erosion with RUSLE2. In: Proceedings of the Federal Interagency Sedimentation Conference, June 27-July 1, Las Vegas, Nevada. 2010 CDROM.
Interpretive Summary: The Revised Universal Soil Loss Equation version 2 (RUSLE2) is the most recent of a family of models proven to provide robust estimates of average annual soil erosion caused by rainfall from a wide range of land use, soil, and climatic conditions. The kind of erosion RUSLE2 estimates is called sheet and rill erosion and occurs from raindrop splash and runoff coalescence into small channels (rills) that occur randomly on the hillslope. RUSLE2 currently cannot estimate the erosion that occurs where surface runoff becomes concentrated within topographic swales within upland fields. Such erosion, called channel or ephemeral gully erosion, is considered by many to be a serious as a threat to soil resources and the environment as sheet and rill erosion. In this report we evaluate a scheme for predicting a series of representative runoff events whose sizes, durations, and timing are estimated from RUSLE2 databases, we link the results to the channel erosion model used in the Water Erosion Prediction Project (WEPP), and compare the results to the ephemeral gully erosion predicted by another model, the Annualized version of the AGricultural Non-Point Source Pollution Model (AnnAGNPS). We also conducted a sensitivity analysis to determine the effect within RUSLE2 of making the largest annual runoff event larger while reducing the number of runoff events to keep total annual runoff the same. Results showed that the RUSLE2 channel erosion estimate did not vary with the size of the maximum runoff event if that event was at least as large as the event that would be expected to occur no more than once per year. However, the channel erosion resulting from the RUSLE2 procedures was larger than that resulting from AnnAGNPS, and the response of the two models to variation in depth to a non-erodible soil layer differed considerably. We conclude that the ephemeral gully erosion process may be of major importance but our confidence in the available state-of-the art tools is low. There is a pressing need for more quantitative measurements of ephemeral gully erosion so that the available models can be validated and improved.
Recent enhancements to RUSLE2 allow determination of a representative runoff event sequence that is intended for calculating concentrated flow erosion. The approach relies on creating estimates of monthly runoff, the number of runoff events per year, and the scale parameter of a gamma distribution describing runoff events for any RUSLE2 climate, soil, and management combination. A sequence of runoff events that totals expected average annual runoff is calculated based on the return period of the largest storm in the sequence, which is assumed to occur on the day of the year with the maximum runoff as determined by disaggregating monthly runoff estimates to daily values. The period between runoff events in the sequence is calculated from the ratio of the maximum event to the maximum daily disaggregated value. A sensitivity analysis was conducted to determine the effect of variation in the return period of the maximum annual runoff event on erosion computed when the resulting storm sequence was coupled with the channel erosion equations of the CREAMS model. We hypothesized that for management systems with more than one tillage operation, there would be at least one maximum in predicted average annual channel erosion as the return period of the maximum storm was increased and the total number of events per year decreased. The channel erosion resulting from the proposed RUSLE2 representative runoff event sequence was then compared with that resulting when 30-year stochastic runoff event populations predicted by AnnAGNPS for the same combination of climate, soil, and management descriptions were coupled with (1) the channel erosion component within CREAMS or (2) the tillage-induced ephemeral gully erosion model (TIEGEM) implemented within AnnAGNPS. The responses of these two channel erosion models to variations in soil erodibility, critical shear stress, and depth of a non-erodible layer were explored.