Optimization of direct bedload measurements using ADCP Bottom Tracking

Cross-channel variability in bedload transport is a predominant phenomenon in gravel-bed rivers, and is attributed to various aspects, including flow conditions, variations in grain-size distribution, boundary shear stress and channel morphology. These variations need to be considered during direct bedload measurements such that the entirety of collected samples is representative of the transport pattern. As a result, the measurement strategies developed and implemented over decades involve sampling at several positions at a cross-section. The distribution of the sampling points across the channel in large rivers has been implemented in various ways: 1) A given number of sampling points distributed at equal intervals along the channel cross-section; 2) One sampling point located on each transport lane; 3) A-priori approaches which allow for evaluations based on the degree of cross-channel variability of bedload transport.

The first approach is still prone to uncertainties to some degree, since it is still unknown whether transport rates in between two sampling locations can produce significant difference in bedload estimations. The second approach is limited to cross-sections where transport patterns are well known and probably not prone to changes. In addition, it would still be uncertain whether any further variations on the transport lanes may be present. Despite considering cross-channel variability, the third method is difficult to implement when bedload is conveyed through a very small section of the river width, since in such case, the method can lead to overly-numerous sampling points that can be relatively difficult to implement in a measurement campaign.

ADCP Bottom Tracking (BT) is an indirect bedload measurement method that utilizes acoustics to detect movement of bed material. At a given point in time, ADCP sensors record properties of acoustic signals emitted and reflected off the mobile bed.  Bedload transport rates are derived from the BT signal using various approaches described by (Conevski, Winterscheid, Ruther, Guerrero, & Rennie, 2018). The continuous recording of an entire cross-section allows the identification of significant variations in transport and hence the derivation of transport lanes and the effective bedload transport width. This method is still under research but its capability to acquire continuous measurements in high-resolution can be harnessed and used to optimize direct sampling.

The current research proposes to complement direct bedload measurements using ADCP-BT measurements, such that the measurements obtained using the latter approach will be utilized in-situ in a-priori assessment of cross-channel variations in transport. The assessment can then be used to adapt the direct sampling strategy. An approach to auto-detect the “appropriate” sampling locations will be developed with the aim to optimally allocate only few sampling points while retaining the original shape of the bedload curve from ADCP-BT measurements. This approach has the potential to reduce uncertainties in the measurements and also provide the possibility of only sampling at sections that are relevant for bedload calculations and thus providing a time-efficient measurement strategy.