Modeling soil erosion and river sediment yield for an intermountain drainage basin of the Central Apennines, Italy

• Soil erosion and river load were analyzed in an Italian Apennine intermountain basin.
• We observed that such regions may have very high values for both factors.
• A sediment yield of 1.33 M t− 1 yr− 1 with a SSY of 32 t haˉ1 yrˉ1 was measured.
• RUSLE outcomes show how sizable sectors of the basin are prone to soil erosion.
• We archived useful elements to enhance our understanding of SSY in these areas.

The overall aim of this research was to investigate the spatial patterns of the soil erosion risk. We focused on accelerated soil erosion processes in an Italian central Apennine intermountain watershed using modeling techniques implemented in a GIS environment. Our thorough literature review revealed a gap in research on soil erosion processes in such forested, intermountain watersheds. To gain a better understanding of the soil erosion processes in such landscapes, we proposed an integrated modeling approach applying a RUSLE model and a Turbidity Unit Index. The model outcomes were validated through measurements of lake sediment deposition. Our findings indicate a potential high soil erosion risk. With 1.33 M t− 1 yr− 1 of annual sediment yield, corresponding to an area-specific sediment yield of 32.35 t ha− 1 yr− 1, the Turano drainage basin belongs to the Italian basins with the highest sediment discharge. The outcomes of the RUSLE model showed that, despite the diverse forms of forests that cover about 62% of the drainage basin area, sizable plots of the investigated area are prone to soil erosion. The validation of the model outcomes revealed that the TU Index model performed significantly better than the RUSLE model with regard to sediment yield prediction. Accordingly, we found that even though rill and interrill processes reach very alarming values (RUSLE), they are not the dominant sediment source within the Turano watershed. Other geomorphological processes contributing to the watershed sediment yield – for instance, megarill, gully, bank and channel erosion and re-entrainment of landslide sediments – were very active in the study area. If both models are used in a combined approach, the amount of river load (TU Index) as well as the relative spatial distribution of rill and interrill erosion processes (RUSLE) can be described with sufficient precision.