The research executed within the framework of this PhD was aimed at determining the contributions of different sediment sources within the Nete and Demer tributary basins of the river Scheldt basin (Belgium).To gain this insight 3 separate topics were researched.
Determining the total suspended sediment flux
The total sediment flux was determined for monitoring locations on the Kleine Nete (Grobbendonk), Demer (Aarschot), Gete (Halen) and Mangelbeek (Lummen), using both measurement data and rating curve estimates.
Additionally, a detailed investigation into the values and timing (hysteresis) of sediment concentrations observed in the Demer at the Aarschot measurement location showed that the river system underwent some significant changes in sediment supply transitioning from an originally slightly sediment-depleted to a sediment-enriched system, caused by impactful maintenance works executed from November 2007 to March 2009. Subsequently, the river system has been gradually returning to its slightly sediment-depleted state.
Budgeting authigenic sediment contributions
In fluvial systems, authigenic sediment is created when groundwater, laden with solutes, seeps into the surface water, where it is subjected to different reigning environmental conditions. As such, authigenic sediment can be a substantial source of sediment, however one which is often overlooked in sediment flux studies.
In case of the Nete basin and the northern part of the Demer basin, ferric authigenic sediment is a major contributor to the total sediment load of the rivers, as can be visually confirmed by the rusty colour of the rivers.
The annual contribution of the ferric authigenic sediment to the total sediment load measured at the Grobbendonk measurement location on the Kleine Nete was both theoretically deduced, as well as modelled, using MARS (Model for Authigenic River Sediment).
The modelling results placed the average annual contribution for the decade 1999-2009 at 61%.
Composite Sediment Fingerprinting
In the final part of this PhD research, the internationally applied composite sediment fingerprinting approach was assessed for its applicability in the Demer basin. The methodology was used to determine the contributions of eight principal tributaries to the total load observed in the Aarschot monitoring location on the Demer.
Several composite fingerprints were created, comprised of different combinations of properties, which had been determined using XRF-analysis. Even though every composite fingerprint was sufficiently capable of discriminating between the eight tributaries, these fingerprints all struggled to generate environmentally realistic results. Different combinations of correction factors (grain size, Loss on Ignition and discriminatory weighting correction factor or omitting all correction factors) have been applied and investigated in more detail to determine their impact on the modelling result.
The reasons behind the poor performance of the sediment fingerprinting approach can most likely be found either in the modelling set-up, in human intervention, causing un-sampled sources to enter the river system, distorting the sediment fingerprint, or in incorrect grain size and organic matter corrections, distorting the fingerprinting signals. A more complex grain size correction factor is needed to drive the model, however, this is difficult to determine due to the flocculation present in the authigenic ferric material.
