A new statistical prediction model, aimed to assess the mean annual suspended sediment yield (SY) in ungauged rivers of Italian Apennines, is here presented. A multiple linear regression equation was obtained by the stepwise technique investigating the relationships between a series of hydro-geomorphometric variables and sediment yield data from 41 river basins located along peninsular Italy and Sicily. The model variables were computed from hydrological data records and from vector river network lines and drainage divides derived from official maps and DEM. The analysis revealed a large variance in the observed sediment yield data. Nonetheless, an optimal result in terms of model significance and efficiency (r2adj = 0.91 at p < 0.05 significance level; Nash–Sutcliffe model efficiency NSE = 0.855; Willmott’s indexes of performance: d = 0.965; d1 = 0.862; dr = 0.862) was finally achieved. According to stepwise regression, the catchment relief and perimeter, together with the topological organization of stream network, the bedrock erodibility and the stream gradient, expressed by specific parameters, appeared as determinant features for SY prediction in the catchments here investigated. The developed model could be helpful to practitioners and scholars for rapid assessments of SY at any point of the river network in the geographic area here concerned. However, the same technique can be utilized wherever in the world since the statistical method itself allows to recognize the number and type of parameters to be used in a given geographic area, on the basis of their local significance. © 2021, The Author(s), under exclusive licence to Springer Nature Switzerland AG part of Springer Nature.

cited By 0