From Hype to an Operational Tool: Efforts to Establish a Long-Term Monitoring Protocol of Alluvial Sandbars using ‘Structure-from-Motion’ Photogrammetry

Abstract:

Despite recent advances in the use of ‘Structure-from-Motion’ (SfM) photogrammetry to accurately map landforms, its utility for reliably detecting and monitoring geomorphic change from repeat surveys remains underexplored in fluvial environments. It is unclear how the combination of various image acquisition platforms and techniques, survey scales, vegetation cover, and terrain complexities translate into accuracy and precision metrics for SfM-based construction of digital elevation models (DEMs) of fluvial landforms. Although unmanned aerial vehicles offer the potential to rapidly image large areas, they can be relatively costly, require skilled operators, are vulnerable in adverse weather conditions, and often rely on GPS-positioning to improve their stability. This research details image acquisition techniques for an underrepresented SfM platform: the pole-mounted camera. We highlight image acquisition and post-processing limitations of the SfM method for alluvial sandbars (10s to 100s m²) located in Marble and Grand Canyons in a remote, fluvial landscape with limited field access, strong light gradients, highly variable surface texture and limited ground control. We recommend a pole-based SfM protocol and evaluate it by comparing SfM-derived DEMs against concurrent, total station surveys and TLS derived DEMs. Error models of the sandbar surfaces are developed for a variety of surface characteristics (e.g., bare sand, steep slopes, and areas of shadow). The Geomorphic Change Detection (GCD) Software is used to compare SfM DEMs from before and after the 2014 high flow release from Glen Canyon Dam. Complementing existing total-station based sandbar surveys with potentially more efficient and cost-effective SfM methods will contribute to the understanding of morphodynamic responses of sandbars to high flow releases from Glen Canyon Dam. In addition, the development and implementation of a SfM-based operational protocol for monitoring geomorphic change will provide a methodological foundation for extending the approach to other fluvial environments.