Supplement - Structure from Motion Metadata and Outcomes

Ground control points used to ensure Structure from Motion (SfM) terrain models were georectified (Westoby et al. 2012; Wolf 2021) using Emlid R2 RTK (real-time kinematic) GNSS (global navigation satellite system) system consisting of a base station set up over an established known point (established with Canadian Geodetic Survey of Natural Resources Canada (NRCAN) service Canadian Spatial Reference System Precise Point Positioning (CSRS- PPP)) and a rover. 

Once the ground control points were surveyed, aerial drone images were acquired. We created flight polygons in Drone Deploy. Pictures were captured with a DJI Mavic II drone with minimum 80 % overlap of photos. Drone Deploy was chosen because it has an option to account for the doming error commonly found in models created from drone imagery and structure from motion (SfM). The doming effect is a systematic error that impacts the DEMs vertical component and can provide errors larger than the usual centimeter level (Sanz-Ablanedo et al. 2020). Generally, each site was flown once in fall of 2020 and once in spring of 2021. 

We created orthorectified images and digital terrain models using Agisoft Metashape, a photogrammetric processing software application that uses SfM. We followed the workflow outlined in Bywater-Reyes and Pratt-Sitaula (2022). Once processed, orthorectified imagery and Digital Elevation Models (DEMs) were exported to ArcGIS Pro for additional analysis. Data collection metadata and postprocessing outcomes can be found in this Zenodo repository.