UAS mapping of surface roughness and digital grain size to assess pre-dam removal baseline conditions along the mainstem Klamath River corridor below Iron Gate Dam, California
Surface roughness and grain size in river corridors are fundamental indicators of river hydraulics. In hydraulic models for coarse-grained rivers, the roughness parameter is often assumed to be related to a representative grain diameter. This paper documents a workflow for using aerial imagery and Structure-from-Motion (SfM) photogrammetry to map surface roughness and digital grain size (DGS) on gravel bars. In June 2022, an Uncrewed Aerial System (UAS) was used to collect aerial imagery at 7 study sites distributed longitudinally downstream from Iron Gate Dam along the mainstem Klamath River. The study sites were selected for repeat monitoring as part of a larger effort to monitor post-dam removal response of the Klamath River to a temporary increase in fine-sediment flux. Advantages of SfM, in comparison to other methods used to acquire high-resolution topography data, is the ability to generate topographic data and orthophotos for DGS analysis from imagery easily collected at multiple spatial scales. UAS imagery was collected at three altitudes (62-m, 15-m, and 4-m), and a workflow for creating georeferenced surface roughness and grain size maps at the patch (100 m2) to site-scale (100,000 m2) was developed. The multi-scale UAS imagery was post-processed, and Digital Surface Models (DSMs) and orthophotos were exported, using AgiSoft Metashape Professional and standardized SfM methods. The DSMs were detrended using an automated tools implemented in ArcGIS to remove bedform gradients and the reach slope. Surface roughness maps were created using the standard deviation of detrended elevations. Plots for DGS analysis were randomly selected from the orthophotos, and DGS was estimated using a wavelet-based method, which minimizes surface disturbance and does not require calibration. For the 62-m survey, the grid-cell resolutions for the DSMs and surface roughness maps was 4 cm, and the orthophotos have a resolution of 2 cm. For the 15-m survey, the resolution of the DSMs and surface roughness maps was 8 mm, and the orthophotos have a resolution of 4 mm. For the 4-m survey, the resolution of the DSMs and surface roughness maps was 3 mm, and the orthophotos have a resolution of 2 mm. The surface roughness maps for the 62-m, 15-m, 4-m surveys provided surface roughness maps suitable for assessing variations in roughness heights across the study sites. The 15-m survey also provided orthophotos suitable for heads-up digitizing of coarse- and fine-grained facies. The 4-m survey provided orthophotos with the grains clearly resolved and suitable for DGS. The wavelet-based DGS analysis used in this study is most appropriate for assessing relative differences in the grain size distributions across a single site or among multiple sites, instead of absolute estimates. This workflow produced a series of digital products to support multi-scale investigations at resolutions appropriate for detecting textural changes in grain size on gravel bars and for producing roughness maps to define roughness parameters for hydraulic modeling.