Effects of groundwater development on surface flow erosion during simulated rainstorms in a laboratory flume

Flume experiments were conducted in a 10 m x 0.8 m indoor flume, with a 0.1 m soil depth above a solid flume floor. Two hours of 42-45 mm.h{dollar}sp{lcub}-1{rcub}{dollar} simulated rainfall resulted in groundwater development in all experiments, usually rising to the soil surface before rainfall ended. Ten experiments were conducted with a Pontypool sandy loam/Peel clay mixture soil: three experiments at a 1.5{dollar}spcirc{dollar} gradient, four at 5{dollar}spcirc{dollar}, and three at 9{dollar}spcirc{dollar}. Three additional experiments were conducted at 5{dollar}spcirc{dollar}: one each with Swinton silt loam, Woburn loam, and Brighton sand. Collected data included surface hydraulic conditions and sediment along the flume, weir sediment and flow collection, and groundwater monitoring with micro-standpipes and time domain reflectometry (TDR).; Surface soil erosion rates during rainstorms increased when a water table developed at a small depth (15 cm or less), on most gradients and all but strongly sealing soils. Within each group of experiments erosion rates increased at consistent and predictable times; gradient and soil type had important effects. Headcut formation and rill initiation commonly occurred with water table rise, and coincident increases in erosion rates were caused primarily by water table rise, with headcuts and rills a supplementary effect.; Groundwater appeared to influence erosion during a rainstorm through its effect on soil strength, without the direct entrainment of soil particles by seepage flow. Erosion rate increases occurred before the water table reached the soil surface. Small soil column experiments provided additional data showing decreasing surface shear strength with increasing pore-water pressure. Shear strengths reached minimum values at initial water table rise. Erosion then became transport-limited until the end of most experiments, as shown by simple regression analyses of weir flow discharge and sediment concentration.; There was no unique morphologic expression of the influence of groundwater; surface effects were only detected with subsurface instrumentation. Groundwater development explained much variation in hydraulic conditions at rill initiation. This study provides evidence that groundwater development can be more important than hydraulic parameter values in controlling erosion rates. Groundwater should be monitored in future soil erosion studies.