A reach of the North Santiam River, Oregon, was used as a case study in an ongoing effort to develop and test uncrewed aircraft system (UAS)-based salmon habitat mapping techniques using: (1) particle image velocimetry (PIV) for estimating surface flow velocities from remotely sensed data; and (2) two-dimensional (2D) flow modeling based on remotely sensed topography and bathymetry (topo-bathymetry). Direct measurements of flow velocity were obtained using an acoustic Doppler current profiler (ADCP) and used to assess the accuracy of the image-derived velocity estimates and modeled flow fields. Water depth was measured using a single beam echosounder and was used to calibrate and validate image-derived depth estimates and to test the accuracy of the flow model. The topography of dry land and water surface elevations were measured using UAS-based near-infrared (NIR) light detection and ranging (lidar) data. River bathymetry was mapped by applying a spectrally based depth retreival algorithm to multispectral image data. Video was acquired from a small UAS and used as input to a PIV algorithm.
The in situ velocity measurements were collected using a SonTek M9 RiverSurveyor ADCP deployed from a cataraft. The SonTek RiverSurveyor Live software package was used to set up the ADCP prior to data collection, control the instrument, view the data in real-time, and save the raw data in MATLAB *.mat data files. A total of four passes back and forth across the channel were completed at seven cross sections located within the field of view of the UAS-based videos. These files were then read into the USGS Velocity Mapping Toolbox (VMT) and further processed to combine the four passes into a single mean cross section and compute depth-averaged velocities (Parsons et al., 2013). The VMT output was summarized by creating a single csv file consisting of a header row with variable names and five columns: 1) East_meters: easting (x) spatial coordinate in meters; 2) North_meters: northing (y) spatial coordinate in meters; 3) velU_meters_per_second: east (u) component of the depth averaged velocity vector in meters per second; 4) velV_meters_per_second: north (v) component of the depth averaged velocity vector in meters per second; and 5) velMag_meters_per_second: velocity magnitude in meters per second. The spatial coordinates are in the UTM Zone 10 projection, WGS84 datum.
The depth measurements in this data release were obtained using a single beam echosounder and are provided in a comma-delimited (*.csv) text file with three columns: East_meters, North_meters, Depth_meters; the units of the spatial coordinates and the depths are meters. The spatial coordinates of the depth data are in the UTM Zone 10 projection, WGS84 datum.
NIR lidar data on the North Santiam River were acquired on July 25, 2022, to measure elevations on dry land and the water surface, and to support the development of a 2D flow model. These data were collected using a Qube240 lidar scanner. The Qube240 uses a YellowScan UltraSurveyor lidar scanner integrated with an Applanix 15 inertial navigation system (INS). The data were acquired from a Quantum-Systems Trinity F90+ UAS platform and were used to produce an interpolated topographic raster Digital Elevation Model (DEM's) with a 1 m cell size in GeoTiff format. The map projection and datum for the lidar contained in this data release is UTM Zone 10 N and NAD83.
Multispectral imagery data on the North Santiam River were acquired on July 25, 2022, to map river bathymetry, which is a required input for flow modeling. These data were collected using a MicaSense RedEdge-MX camera, which is integrated with the Trinity F90+ UAS platform. The RedEdge-MX is a radiometrically-calibrated spectral imager with ten bands between 400 and 900 nm. The redEdge-MX multispectral imagery had pixel sizes of 0.085 m at a flying height of 120 m, and an orthoimage is provided in GeoTiff format. The multispectral image was used to map water depth using the Optimal Band Ratio Analysis spectrally based depth retrieval algorithm (Legleiter and Harrison, 2019). The map projection and datum for the multispectral image contained in this data release is UTM Zone 10 N and NAD83.
Following completion of the depth mapping, we subtracted the image-based depth estimates from the lidar water surface to convert depths to bed elevations using the approach implemented in the ORByT software package (Legleiter, 2021). We fused the bathymetry data with the lidar elevations on dry land to make a continuous DEM, with a resolution of 1 m. The hybrid topographic DEM is provided in this data release as a csv file, file with three columns: East_meters, North_meters, Elevation_meters; the units of the spatial coordinates and the elevation are meters. The map projection and datum for the DEM contained in this data release is UTM Zone 10 N and NAD83.
The DEM contained in this data release was used as input to develop a two-dimensional (2D) hydrodynamic model using the Delft3D-Flexible Mesh (Delft3D-FM, 2023.02 release) model developed by Deltares (2024). We used a curvilinear grid with a cell size of 1 m and included a spiral flow parameter, which accounts for the effects of secondary flow induced by streamline curvature. We set the time step to ensure a Courant number less than 0.7, and specified a minimum depth for wetting/drying calculations of 0.05 m. We prescribed an upstream discharge of 25 m3/s and ran steady flow simulations. To account for turbulence in the model, we used a uniform eddy viscosity value of 0.15 m2/s. The flow resistance was defined using a uniform roughness height (ks) which was converted to spatially explicit Chezy C coefficients via the Colebrook-White equation. Additional Delft3D-FM model input values are provided as a supplemental (*.csv) text file.
Seven UAS-based videos were acquired from a DJI Matrice 210 quadcopter equipped with a Zenmuse X4S optical camera on July 25, 2022. The videos were acquired from a nominal flying height of 120 meters above ground level and are provided in their native form, with a frame rate of 30 Hertz and an *.mov file format. The videos were used to estimate surface flow velocities via PIV, as implemented in the TRiVIA software package (Legleiter and Kinzel, 2023).
References cited:
Legleiter, C. J., and Harrison, L. R. (2019). Remote Sensing of River Bathymetry: Evaluating a Range of Sensors, Platforms, and Algorithms on the Upper Sacramento River, California, USA. Water Resources Research, 55(3), 2142–2169.
https://doi.org/10.1029/2018WR023586Legleiter, C. J., and Kinzel, P. J. (2023). The Toolbox for River Velocimetry using Images from Aircraft (TRiVIA). River Research and Applications, 39(8), 1457–1468.
https://doi.org/10.1002/rra.4147Parsons, D. R., Jackson, P. R., Czuba, J. A., Engel, F. L., Rhoads, B. L., Oberg, K. A., Best, J. L., Mueller, D. S., Johnson, K. K., and Riley, J. D. 2013. Velocity Mapping Toolbox, VMT: a processing and visualization suite for moving-vessel ADCP measurements. Earth Surface Processes and Landforms, 38(11), 1244–1260.
https://doi.org/10.1002/esp.3367
