METHODS
Topographic mapping was accomplished using total stations equipped with digital data collectors. Site size and topographic complexity determined the point density needed to form the topographic models. Smaller sites (~2000 m2) typically require 200-400 points and larger sites (~10,000 m2) require 750-2000 points. Points were also collected offshore to depths of approximately 1 m to provide overlapping coverage with the bathymetry survey. Survey protocol was developed during the GCES Phase II test flows [Beus et al., 1992] and documented according to standard survey practices for ground surveying. Benchmark and backsight relationships were verified at all sites in March, 1991. Terrestrial survey coverage typically extends from the 142 m3/s (5,000 ft3/s) stage elevation to above the 1,274 m3/s (45,000 ft3/s) stage elevation.
A hydrographic survey system expanded ground-based coverage to include the entire river channel and recirculation zone of each fan-eddy complex. This system was not deployed during the November 1997 survey trip because hydrographic surveys require a motor-powered boat and the survey was conducted during non-motor season in GCNP. The hydrographic system consists of a shore-based total station, a boat-mounted transducer, a digital/analog receiving unit, and a computer that controls the data collection process. The shore station data was radio-telemetered to the boat computer where depth-position data is calculated and automatically stored. The location of the boat was determined by targeting a reflective prism mounted directly above the transducer. Digital depth records were checked by comparison with the analog sonar recording. Channel and eddy surveys were made by crossing the river at 7.5 to 10 m intervals combined with upstream and downstream longitudinal lines to form a grid
The ground-based and bathymetric survey points were combined and used to form a Triangulated Irregular Network (TIN) surface model of channel, eddy, and sand bar topography . Breaklines were coded during ground-based data collection along identifiable features (ie. cutbanks, water surface lines, slope breaks, etc.). Sand bar volumes from daily repeated surveys at a single bar were within three percent of each other [Beus et al., 1992]. Therefore, we consider sand bar changes greater than three percent as significant. Verification of x,y position and depth data found that hydrographic survey data have a horizontal error of <1 m and z elevation data < 0.5 m. Eddy and channel volumes were rounded off accordingly to reflect these errors.
The topographic surface model of each site was used to generate profiles, comparison maps, and area and volume calculations. To quantify changes in sediment redistribution within the recirculation zone and the main channel, areas and volumes were calculated within boundaries that approximate the dimensions of the recirculation zone and adjacent channel (Figure 1).
A fixed boundary was established between the main channel and recirculation zone by estimating the position of the eddy fence, the streamline dividing downstream flow and the eddy, and by assuming this zone extends vertically to the bed (Figure 1). Eddy fence location was determined by visual observation of oblique daily photographs, by aerial photographs, and by the positions of separation and reattachment points surveyed in the field at different discharges. This general approximation of eddy fence location best represents the dimensions of the recirculation zone at flows between 566 and 1,274 m3/s.
Areas of deposition and erosion in the recirculation zone were calculated above different topographic levels (Figure 1). An empirically derived stage-discharge relation determined by Kaplinski et al. [1995] at each site and used to define the elevation range of specific flows. The 142 m3/s stage elevation is the minimum discharge at which dam releases occur and was used to separate calculations of bar changes from changes in the deeper, continuously inundated portion of the recirculation zone, which we term the "eddy", and the main channel. We use the term "sand bar" for that part of the recirculation zone above the 142 m3/s stage elevation. Volume and area changes determined within the bar boundary between 142 m3/s, 283 m3/s, 425 m3/s, 566 m3/s and above the 566 m3/s stage elevation contours. The use the term "high elevation" for sand bar changes that occur above the 566 m3/s stage elevation. The 566 m3/s discharge level was chosen to define high-elevation sand bars, because this discharge level was the highest operating limit for Glen Canyon Dam in the 1990's under the interim operating criteria. In addition, this was also the discharge above which bars were considered campsites by Kearsley and Quataroli [1996] and Thompson et al. [1997]. Downstream from the reattachment point of the eddy fence boundary, the 142 m3/s stage elevation was used as the fixed boundary to define the downstream extent of the bar and to separate calculations of main channel change from changes of the bar.
The size of each recirculation zone differs. To compare sites of different size, volume and area are expressed as the percent change from one survey to the next, and as a percentage of the pre-flood surveys conducted in February, 1996. Areas are only reported for sand bars. Because of boundary conditions imposed by the bedrock or talus-confined channel, eddy and channel computational areas do not change appreciably. Changes in elevation of the channel bed are assumed to result from the removal or addition of sand-size or smaller sediment.
Topographic surveys determine the amount of sediment stored on sand bars.
hydrographic surveys determine the amount of sand stored in the channel.
Figure 1. Schematic diagram showing volume calculation boundaries and flow levels.