Description: Underwater photographs were acquired on the Upper Columbia River (UCR) in October-November 2018 and July 2019 in support of the Upper Columbia Remedial Investigation and Feasibility Study (RI/FS). The field survey methods are detailed in the Field Survey Report (FSR). Underwater photographs were acquired at predefined point locations (stations) positioned along predefined river transects. Following data acquisition, the underwater images were processed using Adobe Lightroom and ImageJ software for quantitative, semi-automated sediment grain size analysis. Statistics from each underwater image were calculated to produce a percent composition for each sediment grain size classification, including fine-grained sediment (grain size <2 millimeters (mm) including sand, mud, silt, and clay), gravel (grain size 2 - 64 mm), cobbles and boulders (grain size > 64 mm) and bedrock. The image locations and interpreted sediment composition data were exported to an ESRI shapefile. The data processing methods are detailed in the Data Summary Report (DSR).
Description: The Bureau of Reclamation's (Reclamation) Sedimentation and River Hydraulics (Sedimentation) Group of the Technical Service Center conducted the 2010-11 Franklin D. Roosevelt Reservoir (Lake Roosevelt) bathymetric survey and developed the reservoir topography. A summary of the survey and analysis can be found in Technical Report No. SRH-2012-06 titled “Franklin D. Roosevelt Lake – Grand Coulee Dam 2010-11 Survey.” The survey of Lake Roosevelt, requested by the Reclamation’s Pacific Northwest Region Office, began in October 2010 and continued during portions of June, July, and August of 2011. The 2010-11 bathymetric data was the first detailed bottom information of the deeper portions of Lake Roosevelt. The primary purpose was to develop detailed digital reservoir topography below elevation 1,160 (project vertical datum). A 1974 aerial survey developed 10-foot contours from elevation 1,160 to 1,290 (project vertical elevations). This study measured the project vertical datum (1947 Supplemental Adjusted Datum) to be around 2.5 feet less than the North American Vertical Datum of 1988 (NAVD88). The underwater survey was conducted between water surface elevation 1,260 and spilling elevation 1,290 (project datum in feet). The generated reservoir topography elevations by this study were tied to NAVD88 and horizontal datum was based on Washington’s state plane coordinate system, north zone, in North American Datum of 1983 (NAD83) in feet. The 2010-11 bathymetric survey used sonic depth recording equipment interfaced with real-time global positioning systems (GPS) that provided continuous sounding positions throughout the underwater portion of the reservoir covered by the survey vessels. A multibeam system was the primary instrument utilized to map the main channel of the reservoir from the dam upstream 134 miles to the Northport boat ramp. Multibeam data was also collected on several side tributaries, such as, 10 miles upstream from the Sanpoil River confluence at Keller Ferry and over 30 miles upstream from the Spokane River confluence. The 1974 contour, elevation 1,200, was used as a guide for the boat during the bathymetric collection to assure adequate data overlap between the 1974 contours and the 2010-11 bathymetric data sets. Depending on the geometry and reservoir water level during collection, the outer beams of the multibeam unit extended above elevation 1,200 for the majority of the reservoir surveyed in 2010-11. A single beam collection system was used to collect underwater data along the shoreline and in shallow water areas not covered by the multibeam system. The single beam system was not used throughout the reservoir since the main focus of the smaller vessel was to provide support for the multibeam vessel. The multibeam and single beam soundings were combined and filtered into multiple 10-foot grid files totaling 22.7 million xyz points representing the majority of the deep channel portions of the reservoir. The reservoir topography was generated by importing digital coverages from several sources including contours developed from 1974 aerial photogrammetry, 2007 bathymetry of the upper reservoir, and 2009-10 Light Detection and Ranging (LiDAR) surveys. These data coverages, combined with the 2010-11 bathymetric data, were used to generate digital topography of Lake Roosevelt. The 2010-11 Lake Roosevelt topography elevations were tied to NAVD88. Following is a summary of the data coverages used by this study. 1974 Reservoir Contours - Photogrammetric contour maps at 10-foot intervals of the entire reservoir area were developed from an aerial survey conducted in 1974 when the reservoir was drawn down over 130 feet during the construction of the Grand Coulee Dam third power plant. In 1998-99, digital shapefiles from 160 photogrammetric contour maps were developed by scanning and on-screen digitizing of the reservoir area. The spatial reference of the shapefiles was tied horizontally to UTM Zone 11, North American Datum (NAD83), in meters and the vertical datum was tied to NAVD29 (Supplemental Adjustment 1947) in feet. The coverage of the entire reservoir provided 10-foot contours for elevations 1,160.0 through 1,290.0. There was no metadata with the file. The digital coverage was shifted to Washington state plane coordinates, north zone, in feet and the elevations shifted upward 2.5 feet to NAVD88, in feet, for 2010-11 topographic development. During the 2010-11 bathymetric field collection, the 1974 contours were used as a background to guide the survey vessel to areas lacking data. The areas of the 1974 contours overlapped by the 2010-11 bathymetry were removed. 2007 Bathymetric Data - In 2007 the Colville Tribe conducted a bathymetric survey of the upper arm of Lake Roosevelt from near Kettle Falls upstream to the Canadian Border. There was no metadata on the 5-foot contour coverage. The elevations were assumed tied to the project vertical datum and were shifted to NAVD88 by adding 2.5 feet for the 2010-11 topographic development. The data was collected in the upper portion of the reservoir and backwater elevations may have been affected by inflows, meaning the vertical shift there may be greater. However, insufficient information exists to justify applying a greater shift and it is also possible that the 2007 study accounted for the backwater affect. The areas of the 2007 survey overlapped by the 2010-11 bathymetry were removed. 2009-10 LiDAR Survey - The 2009-10 Columbia River LiDAR survey was a U.S. Army Corps of Engineers (USACE) Northwestern Division, collaborative effort to develop high density terrain data along the Columbia River. The LiDAR data covering Lake Roosevelt was collected between November 2009 and March of 2010 near reservoir elevation 1,280. The survey covered the majority of the Columbia River reach of Lake Roosevelt, but only a small portion of the Spokane River. The data sets that covered Lake Roosevelt were multiple raster and xyz point files. Due to the size of the point files, contours were generated from the raster files and used as hard breaklines for the 2010-11 topographic development. To obtain the detail of the reservoir area, below elevation 1,330, the contours were developed in 1-meter increments starting at the reservoir water surface at the time of the flight. Minor editing was required to remove stray contour lines near the reservoir water surface. For the areas not covered by the LiDAR, the 1974 contours and most recent USGS contours were used to enclose the reservoir. A second coverage from the LiDAR raster files was generated at 5-foot increments from elevation 1,285 to elevation 1,700 (NAVD88) and used during the total reservoir topographic development. For portion of the reservoir, covered by the LiDAR, the 1974 contours of elevation 1,280 and 1,290 were not used. For the reservoir areas not covered by the LiDAR, such as the Sanipoil River across from Keller Ferry and the Spokane River, the upper contours of the 1974 data were used. The 2010-11 generated topographic contours of Lake Roosevelt was developed from combined data sets that included the 2010-11 bathymetry xyz data points and the 2009-10 LiDAR contours as the primary coverages. The 2009-10 LiDAR was flown between reservoir elevations 1,280 and 1,285 (NAVD88) and the 2010-11 bathymetric data mainly concentrated on the deep channel portions of the reservoir below elevation 1,200. Most areas the bathymetric data extended above elevation 1,200 and in vertical bank areas approached the LiDAR coverage. These two coverages did not cover the entire reservoir area and were supplemented with contours from the 1974 aerial survey of the entire reservoir, the 2007 bathymetric survey in the very upper portion of the reservoir, and very small areas some digitized USGS quad contours. The 2010-11 bathymetric data was in a xyz point format, all other coverages were contours and used as hard breaklines for the topographic development. For reservoir topographic development all data coverage layers were combined as Washington state plane coordinates, north zone, in NAD83 and elevations tied to NAVD88. During analysis it was found that in most cases the data sets lined up very well and any differences appeared to be in areas of noticeable shoreline erosion. The portions of the 1974 and 2007 developed contours that were overlapped by the 2010-11 bathymetric and 2009-10 LiDAR data were removed. The 1974 contours of elevation 1,282.5 and 1,292.5 (NAVD88) were removed for all areas of the reservoir shoreline covered by the LiDAR survey. A large portion of the shoreline on the Spokane arm was not covered by the LiDAR and in these areas the 1974 contours were used. The reservoir topography was first generated and checked by dividing the reservoir into multiple sections where a triangulated irregular network (TIN) was developed for each section. From the generated TIN images any data issues could be more easily identified. The TINs illustrated the details obtained from the multibeam data where bottom material and local scour holes were mapped in the main channel. Once any identified issues were resolved, the coverage layers were processed as a terrain using Arc tools. A terrain is based on the TIN method that uses geodatabase feature classes as the data sources. The terrain dataset is a series of TINs that can be quickly displayed as a coarser-grained map within a larger extent, but become finer as the coverage is zoomed to view detailed images. The boundary of the study area and all the file coverages were enclosed by a polygon clip that ran along the alignment of Grand Coulee Dam and assigned no elevation. The terrain took hours to process and resulted in multiple issues when developing contours and computing reservoir surface areas and volumes. The issues were attributed to the large size of the combined sets, so the reservoir was divided into multiple terrains. Dividing the reservoir area into multiple, smaller terrains resolved the contouring issue in all but the upper 30 mile portion of the reservoir using the developed LiDAR contours above the maximum flood pool elevation 1,321.8. This issue was eventually resolved by removing the LiDAR contours above elevation 1,330 for this section of the reservoir only, but other issues arose when attempts were made to compute surface areas and volumes from the terrain near the dam. Some time was spent to resolve the issue, but with no success. The general conclusion was that the problem was likely due to the large file size and detailed reservoir topography in that area. Using Arc tools the developed terrain sections were converted to raster coverages with 10-foot grid details. A raster is a set of cells arranged in rows and columns to represent geographic information and it greatly streamlines the analysis process. The raster coverages were merged into one coverage for the entire reservoir from which surface area and volumes were computed. Contours were also developed from the total raster of the reservoir and compared to the contours from the developed terrains and TINs. Generally, everything matched up well. There were some cases however, such as the very small contours within the mapped deep scour holes, where the raster information appeared not to capture the deeper details. This had little to no effect on the overall volume computations and only occurred in a few cases. For presentation and analysis purposes a hillshade image was developed from the raster. The hillshade is a shade relief from the surface raster that allows visualization of the bottom details in the reservoir areas mapped. For presentation purposes of coverage contours_5ft, the contours greater than 1330 and contour line lengths less than 50 feet were removed. Horizontal datum based on Washington’s state plane coordinate system, north zone, NAD83. The elevations are all tied to NAVD88. Shift the vertical elevations by (-) 2.5 feet to match the Bureau of Reclamation’s construction vertical datum. All elevation shifts were rounded off to the nearest tenth of a foot for this analysis and report. Please refer to technical report no. SRH-2012-06: “Franklin D. Roosevelt Lake – Grand Coulee Dam 2010-11 Survey” for more details.