Exploring Remote Sensing Methodologies for River Bed Grain Size: Insights from a Mountainous Watershed Study in Val Camonica, Italy

The measurement of river bed grain size has become an integral aspect of fieldwork in river geomorphology and regional ecology. Over the past years, various authors have proposed remote sensing methodologies to assess grain size based on ground and aerial images. With the burgeoning applications of small unmanned aerial systems (SUAS) in geomorphology, there is a burgeoning interest in leveraging these remote sensing granulometry methods for SUAS imagery. However, a dearth of studies exists that systematically investigate spatially consecutive images yielding grading curves or specifications over extensive areas within mountainous watersheds.

This study focuses on the granulometry of the mountainous watershed in Val Camonica, located in northern Italy, employing a drone for initial photographic documentation. The study incorporates the BaseGrain software for importing drone spatially consecutive images and extracting granulation curves from the photographed areas. Additionally, the study encompasses the utilization of Structure-from-Motion (SfM) photogrammetry within a Ground Control Points (GCP) workflow to scale the drone-acquired photos. The precision of this scaling is systematically validated by comparing photos with scaling images including meter using BaseGrain software. The precision of AGISOFT software, employed in the SfM-photogrammetry process, is also critically evaluated by itself with different numbers of benchmarks.

Results indicate that, despite the non-professional nature of the instrumentation, the acquisition of high-resolution images is feasible. These images enable the generation of Digital Elevation Models (DEMs) with accuracies ranging between 2 and 3 cm, contingent upon the number of ground control points. The granulation curve, extracted through BaseGrain, exhibits acceptable accuracy within meter-scale resolution. This research contributes valuable insights into the potential of SUAS-based remote sensing granulometry for mountainous watersheds and underscores the importance of methodological precision for reliable results in river geomorphology studies.