Differences between the source contribution of bed material and suspended sediments in a mountainous agricultural catchment of western Iran

• Two types of fluvial sediments were used to analyze sediment contribution.
• The accuracy of two different mixing models was assessed using GOF test.
• Findings reveal the importance of channel bank source.
• Hughes mixing model showed a more accurate results.
• Uncertainty needs to be accounted for sediment fingerprinting.

Soil erosion is intense in semi-arid regions of Iran and causes a decline in dam reservoir capacities and losing fertile soils from agricultural areas. Effective control of sediment delivery to the water storage requires an understanding of the sediment sources. To investigate the spatial provenance of suspended and bed material sediments in the Taleghani catchment, western Iran, eleven geochemical tracers were used to distinguish sediment sources. In total, 44 source samples were collected from the surface soil of three land use sources and the sub-soil of channel banks together with eight suspended samples in different flood events, and eleven bed material samples from different river stream sites were collected. Two mixing models (i.e., the Collins and Hughes mixing models) were applied to compute the contribution of different sources to both river bed and suspended sediments. The results of more accurate mixing model, the Hughes model based on goodness of fit test, indicated that channel-derived sediments dominate (average ~ 71.5%) the sediment sources contributing to bed materials. In addition, temporal variability of sources when suspended load was applied as the sediment input in mixing models showed that crop fields were the dominant source of sediment in flood events with high sediment concentrations. However, in flood events with less sediment concentrations, channel banks reach to higher apportionment.