The erosion phenomenon is very complex and the numerous hydrological subpro-cesses (infiltration, runoff, sediment transport) make very difficult an accurate physically based model.
The existing difficulties of the physically based models, such as the numerous input parameters, differences between the scale of measurements of the input parameters and the scale of the discretization, and uncertainties in the selected model equations, increase the attractiveness of a parametric, even if empirically derived, soil erosion model such as USLE or RUSLE.
For a physically based model, the knowledge uncertainty, due to the incomplete understanding (equations and parameters used) or inadequate measurement or estimate of system properties, and the stochastic variability (temporal and spatial), which is due to random variability of the natural environment studied and is a property of the natural system, can be so high that it jeopardizes the quality of the estimate.
Actually, the physically based model is the future and needs further theoretical deepening of the basic equations and high-quality experimental data, at different temporal and spatial scales, for testing its applicability.
Contemporaneously, simplified mathematical models, such as RUSLE, could be improved to allow a stochastic representation of annual soil loss. The soil loss value and/or sediment yield of given return period is more useful for decision making than a single estimate of the mean annual value that is influenced strongly by extreme values. A probabilistic estimate allows for management based on the level of risk acceptable to resource managers.
Hopefully, high-quality measurements of soil erosion and sediment yield presuppose the definition of standard measurement procedures which have to be established at world scale and that necessarily vary with the spatial (plot, field size, watershed) and temporal (event, monthly, annual) scale.
Finally, the spread of distributed or semidistributed modeling, coupled with GIS employment, needs further developments of soil erosion distributed measurement techniques.
Was this article helpful?