Identifying areas of a basin that are most sensitive to erosion have stimul
ated the study of within-basin variability of the sediment-delivery process
es and the use of spatially distributed models. To verify the reliability o
f a sediment-delivery distributed model applicable at the morphological uni
t scale (i.e., the area of clearly defined aspect, length, and steepness),
experiments were carried out at mean annual and event scales in a small Sic
ilian basin. A Geographical Information System is briefly presented into wh
ich the measurements carried out at the basin outlet (runoff, sediment yiel
d, etc.) and other point and areal information (soil erodibility, digital t
errain model, etc.) were entered.
For validating the model at mean annual temporal scale, the sediment yield
spatial distribution calculated by the model was compared with the net soil
erosion spatial distribution obtained by cesium-137 activity measurements.
At morphological unit scale, the agreement between measured and calculated
sediment yield values showed a good predictive ability of the model at mea
n annual temporal scale.
Finally, the model was calibrated and tested using five rainfall-runoff eve
nts measured at the outlet of the experimental basin. The analysis showed t
hat the coefficient beta, appearing in the expression of the sediment deliv
ery ratio of each morphological unit, is independent of subdivision criteri
on and can be estimated by the hydrological characteristics of the rainfall
-runoff event. The comparison between measured and calculated sediment yiel
d values showed that the sediment delivery distributed approach has a good
predictive ability at event scale, too.