Splash drift on light sandy soils: a field experiment

Vertical splash drift profiles and horizontal splash drift amounts were mea sured during a very wet week on a 7.5 ha agricultural field in Gronheim, Ge rmany. Splash was measured on 45 plots distributed all over the field, at s even (Or eight) heights above the surface. The data show that the vertical splash drift profile is best described by a power function of the form G = a.z(b) (or, alternatively, G = c.z(b) - 1), where G is the total mass of se diment splashed up to a height z and a, b and c are empirical parameters. T he decrease of G with height z is highly variable within the field, even fo r one and the same rain shower. Also, the total horizontal splash drift amo unts show significant differences within the field. There is no clear relat ionship between these splash characteristics on the one hand and the topogr aphic position or the properties of the topsoil on the other. It is suggest ed that spatial variations in rainfall during the rain events are the major cause of the differences in splash drift over the field. Total horizontal splash flux is fairly well predicted by a theoretical splash transport mode l developed earlier by the second author. Due to the effect of the wind (ca using oblique rainfall and an extra-displacement of the soil particles in t he downwind direction) the observed fluxes are somewhat higher than the pre dicted ones, but both are of the same order. The difference indicates the s ignificance of the inclination of the rain with respect to the process of h ydraulic splash.