Abstract
Objectives: The objective of this work is to asses different techniques for evaluating riverbed morphological changes and sediment redistribution in a restored river reach.
Methods: The study was carried out in a restored reach of the Mareiter river, in South Tyrol, Italy. Methods used are (1) repeated cross-section surveys by total station (TS) and (2) Structure from Motion Photogrammetry (SfM) with validation points by RTK dGPS. A DEM of riverbed surface was produced and sediment redistribution was estimated by Difference of DEM (DoD).
Results: Results shows different values of mobilized sediments between the two methodologies. Points restitutions with TS is more accurate than SfM, but spatial resolution is lower, affecting the volumes of mobilized sediments estimation. Errors of TS rise with increasing distances between cross-sections. On the other hand, SfM is not that accurate at point scale, and errors can arise from vegetation cover, shadowsor image reflection. However, their errors do not depend on reach longitude, so the 3D reconstruction model and consequent estimation of sediment redistribution is more accurate than with TS.Advantages of TS is the reliability on riverbed points under water up to depths of wadable conditions, while SfM needs to use a refraction coefficient to correct the bathymetric points for an accurate DEM. Refraction is greater where water is less clear, so typically on riffle, water jetting into pools and high sediment transport rivers (i.e. feed by glacier melt). However, depending on water clarity, SfM bathymetric points can go deeper than wadable depths.SfM involves a relatively fast photo survey in the field, depending on the acquisition method: pole method, RPAS(Remotely Piloted Aircraft System), light weight aircraft, or by helicopter. The first method is even cheaper than the use of TS, while aerial methods are progressively more expansive. SfM can optionally rely on control points to increase accuracy, but they do not need to be permanent like with the TS. Concluding, as new applications (like river restoration monitoring and evaluation) demands higher resolution data, the use of SfM is preferable for estimating morphological changes and sediment redistribution in medium to long reaches, but requires survey integration with other techniques to obtain validation points. Thus, an accurate pre-survey planning is necessary and guidelines for application needed.