Experimental study of surface runoff under simulated rainfall in Cévennes-Vivarais: towards a contributing length.
An experimental study was carried out within the context of OHM-CV with the aim to improve our understanding of runoff generation within watersheds, which is a significant phenomenon during flash flooding. However, specific discharge is well known to decrease with watershed area which implies re-infiltration processes before runoff water reaches the river. Though, re-infiltration is poorly documented in field datasets and thus, difficult to quantify. Using a new rainfall simulator of variable length, runoff was measured at the outlet of parcels of different lengths (1, 2, 4 and 8 m) and for two soil cover types (with and without aerial vegetation). Rain was applied with various intensities and durations. The results confirmed that grassed soil is slower in producing runoff and have a longer recession time than bare soil. No soil cover influence was observed on the quasi-steady runoff regime for a continuous rain. Likewise, no scale effect on this steady regime was observed. To study rain dynamics influence on runoff, non continuous rain application was tested. The results showed that:
(1) The runoff deficit of a non continuous rain for which the flow is interrupted every 10 minutes could reach around 30 % loss as compared to a continuous rainfall of the same amount and maximal intensity.
(2) The mean runoff of a non continuous rainfall alternating 10-minute peaks and voids was approximately equal to the runoff of the continuous rainfall of the same mean intensity.
Finally, vegetated cover seemed to play a role of soil protection against erosion.
We conclude that, rain intermittence and intensity variations might be a dominant factor of re-infiltration within watersheds especially with spatially homogeneous soils.
HyMeX – Hydrological cycle in the Mediterranean Experiment 2010-2020