Grand Canyon Riverbed Sediment Changes, Experimental Release of September 2000 - A Sample Data Set
Florence L. Wong1, Roberto J. Anima1, Peter Galanis1, Jennifer Codianne2, Yu Xia3, Randy Bucciarelli4, and Michael Hamer5
1U.S. Geological Survey, Menlo Park, California 94025; 2Santa Clara Valley Water District, San Jose, California; 3Mason, Bruce & Girard, Inc., Portland, Oregon; 4Scripps Institution of Oceanography, La Jolla, California; 5City of San Leandro, California
|Figure 1. Open floodgates on the Glen Canyon Dam, Arizona, during a previous experimental flood.|
|An experimental water release from the Glen Canyon Dam into the Colorado River above Grand Canyon was conducted in September 2000 by the U.S. Bureau of Reclamation. The U.S. Geological Survey (USGS) conducted sidescan sonar surveys between Glen Canyon Dam (mile -15) and Diamond Creek (mile 220), Arizona (mile designations after Stevens, 1998) to determine the sediment characteristics of the Colorado River bed before and after the release. The first survey (R3-00-GC,
28 Aug to 5 Sep 2000) was conducted before the release when the river
was at its Low Summer Steady Flow (LSSF) of 8,000 cfs. The second
survey (R4-00-GC, 10 to 18 Sep 2000) was conducted immediately after the September 2000 experimental release when the average daily flow was as high as 30,800 cfs as measured below Glen Canyon Dam (Figure
2). Riverbed sediment properties interpreted from the sidescan
sonar images include sediment type and sandwaves; overall changes in
these properties between the two surveys were calculated.
Sidescan sonar data from the USGS surveys were processed for segments of the Colorado River from Glen Canyon Dam (mile -15) to Phantom Ranch (mile 87.7, Figure 3). The surveys targeted pools between rapids that are part of the Grand Canyon Monitoring and Research Center (GCMRC http://www.gcmrc.gov/) physical sciences study.
Maps interpreted from the sidescan sonar images show the distribution of sediment types (bedrock, boulders, pebbles or cobbles, and sand) and the extent of sandwaves for each of the pre- and post-flow surveys. The changes between the two surveys were calculated with spatial arithmetric and had properties of fining, coarsening, erosion, deposition, and the appearance or disappearance of sandwaves.
This report describes GIS spatial data files for this project and provides examples of the data from the Colorado River near mile 2 below the confluence of the Paria and Colorado Rivers. The complete data set includes sidescan sonar images and interpreted map files for each of the pre- and post-flow surveys and the changes between the segments of river listed in Table 1 below.
images were processed with the USGS Mini Image Processing System
(MIPS) <http://terraweb.wr.usgs.gov/software/mips/>, a set of software tools that are used to convert rectangular blocks of raw data to georeferenced images in UTM 12 coordinates in MIPS "raw" or .bil (band interleaved by line) format. The processed images have pixel values 0 to 255, indicating a range of sonar reflectance from low to high. The resolution of the sidescan images is 0.2 meter per pixel. The images were converted to TIFF format for use with illustration software.
TIFF images were imported to Adobe Photoshop where the color mapping
was inverted so that original 0 to 255 pixel values are rendered as
white (0, low reflectance) to black (255, high reflectance) (Figure 4). Bedrock and boulders, which
have high sonar reflectance, image as dark reflectors whereas sand
and gravel, which have low sonar reflectance, image as various shades
of white to dark gray. Features interpreted from the pre- and
post-flow sonar images were drawn as vector layers in Adobe
Vector layers from
Adobe Illustrator files were exported to Arc shapefiles with the
Avenza Mapublisher plugin. Registration tics were generated from the
registration world file of the sidescan sonar images. The upper left
tic coordinates and a scale of 567 supplied to Mapublisher produces
registered shapefiles. The shapefiles were converted to Arc coverages
for editing and tagging.
Because of intermittent GPS coverage in the Grand Canyon, most coverages required further manual registration. The image and vector files in UTM 12 coordinates were converted to Arizona State Plane coordinates for conformance to GCMRC requirements.
Areas of change
were determined with with grid arithmetic. Each of the riverbed types
was assigned a value: bedrock 1, boulders 2, pebbles_cobbles 4, sand
8, and shoreline 16. The pre- and post-survey polygon files were
converted to grids and the R3-00-GC grid was subtracted from the
R4-00-GC grid. The difference values of interest appear in Table 2. A
similar calculation was used to determine areas of "new
sandwaves" and "sandwaves lost".
The Grand Canyon riverbed data are divided into pre- and post-release data sets, R3-00-GC and R4-00-GC, respectively. These data sets are further split into river segments, each of which includes the sonar images and related interpreted vector layers. The R4-00-GC river segments also have polygon files of changes noted between the two surveys. Each river segment, designated by xxx, is represented by the data files and their attributes described in the Table 3 below. The data file inventory appears in Table 4.
The data sets have
been assembled in a Arcview 3.2 project which displays a view of each
of the R3-00-GC, R4-00-GC, and change data sets. A button tool
permits the views to be synchronized to the window extent of any one
of the three views (Figure 5 (76 kB)). For
example, if the extent of the window in view "r400, azsp"
is reset, the extent of the other two views are made to match by
application of the button tool with the following Avenue script:
theView1 = av.GetProject.FindDoc
The segment of the Colorado River mapped in these images is approximately 17 miles downriver from Glen Canyon Dam where an experimental water release was conducted in September 2000. Maps of sediment type and bedforms from before (Figure 6) and after (Figure 7) the water release provide an idea of the the changes (Figure 8) that can be attributed to the increased stream flow during the release. The areas of the polygons of sediment type and bedforms for segment 'bp1a', downstream from where the Colorado River is joined by the Paria River, are provided in Tables 5 and 6 (only about half the segment is displayed in these maps). Within the limitations of the navigation, the calculations provide a qualitative view of the changes on the surface of the riverbed after the flow.
For this segment of the Colorado River near mile 2, areas of bedrock and pebbles or cobbles decreased after the water release. Areas of boulders and sand increased and sandwaves appeared to have decreased. The character of the changes fall into four types: (1) fining if a coarser sediment were replaced by a finer one, (2) coarsening if a finer sediment were replaced by a coarser one, (3) deposition if bedrock became sediment covered, and (4) erosion if sediment were removed to expose bedrock. Areas of fining and coarsening appear in subequal amounts. Areas of deposition are twice erosion for this sample area. Sandwaves disappeared from more areas than appeared in new areas.
Stevens, Larry, 1998, The Colorado River in Grand Canyon: A Comprehensive Guide to its Natural and Human History: Flagstaff, Arizona, Red Lake Books, 116 p.
For more information:
Mary McGann provided a helpful review of an earlier version of this report.
Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
This report is adapted from a poster presented at the 23rd Annual ESRI International User Conference, San Diego, California, July 2003.