Modeling the Reduction in Soil Loss due to Soil Armouring Caused by Rainfall Erosion

Abstract

Surface soil properties can change as a result of soil disturbances, erosion, or deposition. One process that can significantly change surface soil properties is soil armouring, which is the selective removal of finer particles by rill or interrill erosion, leaving an armoured layer of coarser particles which may reduce further soil loss. This phenomenon occurs in a variety of soils with high concentrations of particles larger than 2 mm. Rapid armouring is typically reported in mine sites, construction sites, road embankments, rangelands, and hilly regions with steep and bare slopes. Changes in surface soil properties over time induced by process such as armouring are not accounted for in current soil erosion models such as WEPP or RUSLE, because little is known about rates of armouring over time as a function of rainfall intensity, slopes, and other factors. In this presentation we quantify soil armouring induced by interrill erosion in experiments and propose ways to account for the armouring process in WEPP and RUSLE. Laboratory rainfall simulation experiments were conducted to understand and quantify the armouring process from interrill erosion. Rainfall events with intensities ranging from 22 to 80 mm/h were simulated on 0.56 m2 plots at slopes of 18 and 24 degrees using topsoil from a mine restoration site in New Zealand. Measured soil yields were divided into coarse (>2 mm) and fine particles (<2 mm). Results showed over 75% reduction in total soil loss between freshly applied soils and highly armoured soils at the same slope. Armouring rates for sandy soils and loess type soils are also reported from a series of other experiments using a steep slope flume, as are field observations of armouring in large 25 m long plots in steep reconstructed slopes. The armouring process was modeled with WEPP by changing soil erodibility parameters. In RUSLE, armouring was modeled by altering the percent rock cover over time. Both models performed equally well in predicting yields of fine particles during the armouring process as surface soil properties changed. WEPP was also modified to simulate erosion processes of the coarse particles and results closely matched measurements. It is demonstrated that minor modifications of both models can permit an accurate estimation of interrill erosion during the armouring process and thus improve soil loss estimates; however, interactions between rill erosion and armouring need to be further studied.