Channel responses to the removal of Gold Ray and Savage Rapids Dam. Prepared by Desirée Tullos and Cara Water

Transcription

1 Channel responses to the removal of Gold Ray and Savage Rapids Dam Prepared by Desirée Tullos and Cara Water

2 Introduction. As the results of extensive efforts over several decades, the Rogue River is now free flowing for approximately 189 miles from the ocean to the base of Lost Creek Dam. This project continued monitoring the sediment transport, geomorphic, and habitat changes associated with the removal of Gold Ray Dam (RM 125.7, RK 202.3) on the Rogue River. The intent is to investigate the impacts of removing Gold Ray Dam on the physical conditions of the river in years 3-4 years post removal, and relate those effects to the cumulative results from the recent removal of nearby Gold Hill Dam (RM 121, RK 194.7) and Savage Rapids Dam (RM 107, RK 172.2). Methods: Channel morphology and sediment movement monitoring Field data collection: Data collection by OSU consisted of three components on an annual basis: 1) Boat-based bathymetry of the main channel; 2) ground surveys of bars and sloughs; 3) particle counts on bars (Table 1). The longitudinal extent of the surveys varied by data collection event, with the maximum extent from TouVelle State Recreation Site to Whitehorse Park (Table 1, Figure 1). Boat-based bathymetry was collected from a cataraft with a Teledyne RD Instruments Workhorse Rio Grande ADCP for depth and a Topcon GR-3 RTK GPS, using the ODOT network for a remote base station for water surface elevation and location. The same GPS system was used for the ground surveys. The boat-based surveys consisted of 3-5 longitudinal profiles spread across the width of the channel with additional data collected in alcoves (Figure 2), with the number of profiles varying due to time constraints and equipment issues. The 2012 ground surveys consisted of boat-based surveys of the reservoir (Figure 2) and cross sections in Kelley and Tolo Sloughs. A grid-based sampling was also conducted on bars, including the edge of water delineation. The sloughs were not surveyed in 2013 as they were no longer part of the active river channel. LiDAR and sonar were collected by River Design Group in 2009 from the top of the Gold Ray reservoir to above Nugget Falls were used to fill in edges around these two sections due to large degree of bank failure in the reservoir post-removal and unstable banks downstream. Table 1: Types, dates, and locations of data collection Dates Type of data collection Locations July 23 to Aug. 5, 2010 (Preremoval) Boat-based bathymetry, ground surveys of bars, particle counts of bars TouVelle to Valley of the Rogue (excluding reservoir for ground surveys) Oct. 6-7, 2010 (Post-removal) Ground survey of bars and sloughs, particle counts of bars Former reservoir, sloughs, and immediately downstream June 16-20, 2012 Boat-based bathymetry TouVelle to Baker Park Sept , Oct. Ground survey of bars, sloughs, and TouVelle to Baker Park 12-14, 2012 reservoirs, particle counts of bars July 15-19, 2013 Boat-based bathymetry TouVelle to Baker Park Sept , 2013 Ground survey of bars and sloughs, particle counts of bars TouVelle to Whitehorse Park 2

3 Figure 1: Extent of Channel Morphology (red areas) and Bar Surveys (red and orange areas). Figure 2: Locations of data collected in the reservoir of Gold Ray Dam. Data Processing: For the boat-based bathymetry data, the GPS and ADCP data were combined to derive the channel bottom elevations at the locations sounded by the ADCP using AdMap (Mueller/USGS). Water surface elevations for locations with bad (floating or autonomous) GPS data were determined based upon good (fixed) GPS data upstream and downstream of the locations within the same longitudinal profile, or adjacent longitudinal profiles at the same cross section. For both boat-based and ground-based data collection, locations and elevations of control points were compared between profiles and between surveys to account for differences. Data Analysis: We combined boat-based and ground-based data for the main channel to create a Triangulated Irregular Network (TIN) surface for each annual survey using ArcMap. The TIN surfaces were converted to rasters and compared using the Geomorphic Change Detection (GCD) toolbox within ArcMap (Wheaton et al., 2010). From the GCD analysis, the net change in volume for sections is calculated at lengths of the average channel width (60 m). Cross sections surveyed in the sloughs are plotted for comparison. The particle counts are plotted by cumulative percent finer to determine D16, D50, and D84. Since the locations of bars, and therefore particle counts, changed from between annual surveys, the metrics derived from the 3

4 particle counts (D16, D50, D84) were plotted with respect to distance along the channel centerline. Results From a comparison of the longitudinal profile of the thalweg, there is little difference between the 2010 (pre-removal), 2012 (two years post removal), and 2013 (three year post removal) surveys in the reach upstream of the reservoir (Figure 3) behind Gold Ray Dam. In contrast, the channel immediately upstream of the dam incised up to 5 meters (Figure 3), whereas the bed immediately below GRD aggraded up to 2 meters (Figure 3), with some evidence of pool filling as far downstream as ~14 km (Figure 5). Given the minimal changes between 15 km and 26 km downstream of the Gold Ray Dam, we speculate that channel changes 26 km and farther downstream from Gold Ray Dam are primarily a function of the 2009 removal of Savage Rapids Dam (Figure 5). The first two pools, located within 2.5 km downstream of Savage Rapids Dam, filled between 2010 (one year postremoval for Savage Rapids Dam) and 2012 (three years post-removal for Savage Rapids Dam). Furthermore, there is evidence of additional pool filling up to 3.1 km downstream from Savage Rapids Dam between the 2012 and 2013 surveys (Figure 6). The GCD figures (Figures 11-17) generally confirm changes observed in the longitudinal profile immediately around both GRD and SRD. Between pre-removal and 2012, the reservoir incises up to 4pm, primarily within the first 1.2km upstream of the dam (Figure 12). The differences in reservoir bed elevations between 2012 and 2013 are small (Figure 12). Downstream of GRD, we observed deposition as far downstream as 14km, with maximum depths of up to 4.5, associated with two sand bars located ~10 km and ~ 14 km downstream of GRD (Figure 15). At SRD, GCD figures indicate the reservoir incised as far as 2km upstream from the SRD (16.7 km downstream of GRD), with a maximum aggradation depth of 1.5 m (Figure 16). Downstream of SRD, deposition appears to have occurred as far as 2.7 km (21.4 km downstream of GRD), with a maximum of up to 7 m of deposition (Figures 16, 17). The cross section comparison in the sloughs indicates little change occurred in Tolo Slough (Figures 29 and 30) or in the two most upstream cross sections of Kelley Slough (Figures 27 and 28). However, large changes (1-2 m of cut) occurred between October 2010 (immediately postremoval) and September 2012 for most of the other cross sections in Kelley Slough (Figures 21-26). From the particle counts, bars coarsened far upstream of the Gold Ray Dam from 2010 to 2012, but became finer again between 2012 and Immediately downstream from the Gold Ray Dam, bars became finer from 2010 to 2012 and stayed finer up to 10 km downstream through 2013 (Figures 18-20). However, there were fewer particle counts downstream of the dam in 2012 and 2013, so it is unclear if there is a trend farther downstream. In the area influenced by Savage Rapids Dam, starting 26 km downstream of Gold Ray Dam, there is no discernable trend or difference between particle sizes in 2010 and

5 Figure 3: Longitudinal Profile for TouVelle State Recreation Site to above Nugget Falls Figure 4: Mean Change in Bed Elevation by 60 m Section for TouVelle State Recreation Site to above Nugget Falls 5

6 Figure 5: Longitudinal Profile for below Nugget Falls to Powerhouse Rapid Figure 6: Mean Change in Bed Elevation by 60 m Section for below Nugget Falls to Powerhouse Rapid 6

7 Figure 7: Longitudinal Profile below Powerhouse Rapids to Evans Creek Figure 8: Mean Change in Bed Elevation below Powerhouse Rapids to Valley of the Rogue State Park 7

8 Figure 9: Longitudinal Profile from Evans Creek to Baker Park Figure 10: Mean Change in Bed Elevation from the top of Savage Rapids Reservoir to Baker Park. The black vertical line represents the location of Savage Rapids Dam. 8

9 Figure 11: Change in bed elevation from pre-removal to 2012 and 2013 for TouVelle Recreation Site (-8.48km) to the top of the Gold Ray Reservoir (-4.0km) Figure 12: Change in bed elevation from pre-removal to 2012 and 2013 for the Gold Ray Reservoir (-3.3km to 0km) 9

10 Figure 13: Change in bed elevation from pre-removal to 2012 and 2013 for Downstream of the Gold Ray Reservoir (-.12km) to above Nugget Falls (+5.0km) Figure 14: Change in bed elevation from pre-removal to 2012 and 2013 for below Nugget Falls (+6.5km) to above Powerhouse Rapid (+7.8km) 10

11 Figure 15: Change in bed elevation from pre-removal to 2012 and 2013 for below Powerhouse Rapid (+8.9km) to Valley of the Rogue State Park (+17.9km) 11

12 Figure 16: Change in bed elevation from pre-removal to 2012 and 2013 for Savage Rapids Dam former reservoir (+27.7km) to Pierce Riffle Park (+30km) Figure 17: Change in bed elevation from pre-removal to 2012 and 2013 for Chinook Park area (+32.3km to +33km) 12

13 Figure 18. D16 over time around Gold Ray Dam Figure 19. D50 over time around Gold Ray Dam Figure 20. D84 over time around Gold Ray Dam 13

14 Figure 21 Figure 22 Figure 23 Figure 24 14

15 Figure 25 Figure 26 Figure 27 Figure 28 15

16 Figure 29 Figure 30 16