Use of high-resolution imagery acquired from an unmanned aircraft system for fluvial mapping and estimating water-surface velocity in rivers

The use of unmanned aircraft systems (UAS) for environmental monitoring in the United States is anticipated to increase in the coming years as the Federal Aviation Administration (FAA) further develops guidelines to permit their integration into the National Airspace System. The U.S. Geological Survey's (USGS) National Unmanned Aircraft Systems Project Office routinely obtains Certificates of Authorization from the FAA for utilizing UAS technology for a variety of natural resource applications for the U.S. Department of the Interior (DOI). We evaluated the use of a small UAS along two reaches of the Platte River near Overton Nebraska, USA, to determine the accuracy of the system for mapping the extent and elevation of emergent sandbars and to test the ability of a hovering UAS to identify and track tracers to estimate water-surface velocity. The UAS used in our study is the Honeywell Tarantula Hawk RQ16 (T-Hawk), developed for the U.S. Army as a reconnaissance and surveillance platform. The T-Hawk has been recently modified by USGS, and certified for airworthiness by the DOI - Office of Aviation Services, to accommodate a higher-resolution imaging payload than was originally deployed with the system. The T-Hawk is currently outfitted with a Canon PowerShot SX230 HS with a 12.1 megapixel resolution and intervalometer to record images at a user defined time step. To increase the accuracy of photogrammetric products, orthoimagery and DEMs using structure-from-motion (SFM) software, we utilized ground control points in the study reaches and acquired imagery using flight lines at various altitudes (200-400 feet above ground level) and oriented both parallel and perpendicular to the river. Our results show that the mean error in the elevations derived from SFM in the upstream reach was 17 centimeters and horizontal accuracy was 6 centimeters when compared to 4 randomly distributed targets surveyed on emergent sandbars. In addition to the targets, multiple transects were surveyed in the reaches for detailed cross-section comparisons to the photogrammetric surface model. When operating the T-Hawk in a hover and stare flight pattern, natural tracers (floating algae) were successfully detected for use in estimating water-surface velocity. A subsequent flight with the T-Hawk is planned this fall to evaluate the ability of the UAS system to detect geomorphic change between successive surveys. While the vertical accuracies and the spatial extent of the T-Hawk system are currently less than conventional airborne LiDAR mapping, further technical advancements could increase the accuracy of UAS products providing a relatively low-cost solution for monitoring project-scale river management activities at a high temporal resolution.