Analysis of the relationship between soil loss and soil roughness at the Masse experimental station (central Italy)

The relationship between the soil roughness and the event soil loss on bare soil conditions has been investigated at the Masse experimental station (central Italy). The analysis was first performed on two micro-plots (width 1 m, length 0.92 m, slope 16%) by means of repeated rainfall simulation experiments. Each experiment started after tillage and included three subsequent rainfall events carried out in the range of few days. Rainfall events had an average constant intensity of about 68 mm/h, duration between 60 and 105 min, and were preceded by a wetting phase (with no runoff). A positive correlation between soil loss (or sediment concentration) and soil roughness was found for low-runoff conditions. Probably in these circumstances, the sediment detachment is mainly caused by the splash effect, whose effectiveness increases with the roughness. Instead, for high-runoff conditions (i.e runoff coefficient > 0.8) the correlation becomes decisively negative. In this case the sediment detachment and transport are controlled by the runoff whose effectiveness decreases with roughness. A similar correlation analysis was then carried out considering the event soil loss database collected for the Wischmeier plots of the experimental site since 2008 (629 records). The sign of the correlation was confirmed, but its magnitude is much lower than that observed in the more controlled microplot experiments. In conclusion, the analysis showed that soil roughness is an important driver of the event soil loss and that its effect varies with the runoff, i.e. with the characteristics of the event. However, a profitable use of a roughness variable (or information) to improve the present USLE-derived models at event and plot scale appears difficult, due to uncertainty issues, and due to the prevailing effect of runoff in the events with major soil losses.