Original paper

Infiltration rates affected by land levelling measures in the Souss valley, South Morocco

Peter, K.D.; Ries, J.B.

Abstract

For the investigation of changes in infiltration rates by land levelling measures, measurements by means of simulated rainfall and cylinder infiltrometer were operated on an ample alluvial fan in the Souss valley, South Morocco. Field data was collected on both levelled and non-levelled test sites. Infiltration rates were very low on levelled test sites, where soils are sealed by crusting. Higher infiltration rates were found on non-levelled test sites, with less crusting and more vegetation cover. Both rainfall simulation as well as cylinder infiltrometer are suitable to study the infiltration process. Due to the small observation area encompassed for each experiment, the latter needs more replications. Measurements by rainfall simulation experiments are more elaborate, but also more realistic and accurate. The infiltration rates measured by means of ponding were 3.8 times higher than by means of simulated rainfall at an intensity of 40 mm h-1. Both methods result in lower median infiltration rates on levelled test sites, 16 mm h-1 for rainfall simulations and 30 mm h-1 for ponding. On non-levelled test sites infiltration rates are 1.5 times higher for rainfall simulations and respectively 2.6 times higher for ponding. This means there is more runoff water on areas with land levelling measures which can lead to a higher potential of erosional hazards. Other influential factors for infiltration are crusting and grain size distribution. Furthermore, a correction factor was found to derive the actual infiltration rate out of potential infiltration rates by means of ponding. This is only applicable for consistent rainfall intensities and homogeneous test areas.

Keywords

infiltrationland levellingsoil erosionmorocco