INTEGRATED ASSESSMENT AND ANALYSIS OF PHYSICAL LANDSCAPE PROCESSES THAT IMPACT THE QUALITY AND MANAGEMENT OF AGRICULTURAL WATERSHEDS
Location: Watershed Physical Processes Research Unit
Title: Predicting Ephemeral Gully Erosion with RUSLE2
Submitted to: National Sedimentaton Laboratory (NSL)- 50 Years of Soil & Water Research in a Changing Agricultural Environment
Publication Type: Proceedings
Publication Acceptance Date: September 3, 2008
Publication Date: N/A
Interpretive Summary: RUSLE2 is the erosion model that the Natural Resources Conservation Service (NRCS) uses to estimate soil erosion on farmland and to determine farm program eligibility. RUSLE2 does a good job of estimating erosion on hillsides, but it cannot estimate erosion in concentrated flow channels like the “ephemeral gullies” that form in cropped fields during intense storms that the farmer smooths over the next time he tills his field. Adding the ability to account for this kind of soil erosion has been identified as a priority need by NRCS. In this paper we explore a way that rainfall runoff and ephemeral gully erosion can be estimated using the information already available in the RUSLE2 databases. We compare the runoff predictions with those of more a complex and proven runoff estimation model and with measured runoff and erosion from plots at Holly Springs, MS. The results are encouraging. When the methods presented in this paper are incorporated into RUSLE2, conservation planners will be able to estimate long term average ephemeral gully erosion for any farm field in the continental United States.
RUSLE2 provides robust estimates of average annual sheet and rill erosion from a wide range of land use, soil, and climatic conditions, but it currently cannot calculate ephemeral gully erosion, which capability has been identified as a priority need by user agencies. One obstacle is RUSLE2’s use of long-term daily average precipitation, which essentially applies a small amount of erosive rainfall to every day during the year. This works well for the linearly-scaled rainfall erosivity impact on detachment, but cannot be readily translated into runoff amounts that drive detachment and transport in channel flows. We propose and evaluate a method of predicting the magnitudes of a series of runoff events whose sizes, durations, and timing are estimated from RUSLE2 databases, based on analysis of results of 30-year simulations using a widely-accepted climate generator and runoff model. We then couple the index runoff amounts and rates and the corresponding RUSLE2 hillslope sheet and rill sediment yield estimates with the channel erosion routines used in CREAMS to calculate ephemeral gully erosion. With this procedure, ephemeral gully erosion is be estimated as the sum of a location-specific annual sequence erosion events, whose timing, size, and duration are estimated from the existing RUSLE2 databases and whose total runoff volume approximates local average annual runoff. We illustrate the procedure by applying it to hypothetical 5 ha fields with silt loam soil cropped to cotton (Gossyipium hirsutum, L.) in Marshall County, MS, that are bisected with a potential ephemeral gully having thalweg slopes ranging from 0.5 to 5%. Hillslopes on both sides of the channel were modeled after the erosion plots at Holly Springs, MS, with 5% steepness and 22.1 m length, allowing hillslope predictions to be compared with observations. Results suggest that ephemeral gully erosion may be the same order of magnitude as sheet and rill erosion. Comparison of observed monthly runoff with predicted patterns suggests that the procedures outlined may underestimate plot-scale runoff during periods of the year with greater than average rainfall intensity.