Answers are grounded in this paper and cite their sources.
Soil loss curves of six soil types were obtained using a laboratory rainfall simulator. During the first 10 min of rainfall, soil losses increased to a maximum and then decreased sharply for the Marshall silty clay loam, Dunkirk silt loam, Caribou loam, and Nicholville silt loam soils. There was no peaking effect with the Hartland sandy loam and Winooski silt loam soils and the rate of soil loss was constant after 10 min. A linear relationship was obtained between clay content and soil loss during the 5‐ to 10‐min period. Detailed studies on the Dunkirk and Winooski soils suggest that the peaking effect may be due to the rate and extent of crust formation in the early stages of rainfall wetting. Bulk density of the 0‐ to 5‐mm layer increased from 1.15 to 1.65 g/cc for the Dunkirk soil and from 0.83 to 1.14 g/cc for the Winooski. The amount and distance of splash from a Caribou soil was measured using two drop sizes, 5.1 mm and 3.2 mm. Splash soil losses from the soil pans for the six soils was as much as three to four times those found in the runoff water.
No takes yet. Share an insight, caveat, or question.
Epstein et al. (1967) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: