Monitoring the effects of Marmot Dam Removal, Sandy River, Oregon
In October 2007, Portland Gas and Electric removed the Marmot Dam on Oregon’s Sandy River, liberating 75,000 cubic meters of stored sediment. The dam had blocked upstream passage for salmon and steelhead since 1912. Of the many questions being asked regarding the effects of dam removal, the question of how the sediment would route through the system proved most difficult to answer. Mathematical models were utilized to make predictions of sediment routing and a multi-disciplinary monitoring program was implemented. An interagency effort (involving the USGS, US Bureau of Reclamation, US Forest Service, National Center for Earth Surface Dynamics, Johns Hopkins University, Oregon Watershed Enhancement Board and Oregon State University), was developed to empirically test the models and monitor the sediment pulse. Over the next three years, GMA contributed to two vital components of the study: a volumetric assessment of sediment deposition in several sub-reaches and direct measurement of sediment discharge into the reservoir reach and from the mouth of the Sandy Gorge, roughly five miles downstream of the former dam site.
Topographic-volumetric comparisons following each winter’s high flows were required to test the models and learn more about the routing dynamics of the Sandy River. GMA provided channel bathymetry to dovetail with LiDAR and topographic (total station and RTK GPS) surveys. Unique study reaches were approached differently. Steeper, more confined gorges required crews to hike in with packframes, survey equipment and inflatable kayaks. Islands, scour channels and associated bar features were surveyed for topography using conventional techniques. Class three rapids and deep pools were surveyed using kayaks or advanced swimming and wading techniques. In the depositional reach above the gorge, cross sections were surveyed every 10 meters to dovetail with existing (but incomplete) datasets and provide topography to evaluate changes in storage. The downstream, more alluvial reach was mapped using RTK GPS. A canoe was used to access bar tops, side channels and remote edges of water with a rover. A sixteen foot cataraft, fitted with a depth sounder (interfaced with a rover unit and a laptop computer), was used to map long pools and complex riffles. The highly maneuverable cataraft allowed upstream passes and provided excellent coverage in water as shallow as 1.4 feet.
Measurement of sediment discharge from the mouth of the gorge was accomplished via a cataraft on a cableway in WY2008 and in WY2010. Equipment was hiked in to the remote site and the station was established on the 250 foot wide channel. A streamgage measuring continuous stage was constructed and numerous discharge measurements were collected during floods to provide a continuous discharge record. GMA collected bedload, suspended sediment, water surface slope and turbidity measurements during several large winter storms and one unique snowmelt event. These data were used to create sediment rating curves and were utilized with the continuous flow records to compute continuous bedload and suspended sediment discharge. These load totals were compared to data collected by other researchers to further our understanding the post dam-removal sediment budget along the lower Sandy River.
During WY2009, efforts were shifted to capture sediment transport data coming into the former reservoir reach. GMA constructed a unique trailer-mounted crane system to sample off the 50 foot high bridge near Brightwood. TR2 bedload samplers and D-74 suspended sediment samplers were deployed in this fashion at much higher flows than the USGS was able to sample the previous year. GMA also measured flows of over 11,000 cfs, allowing the USGS to more than double its validated stage-discharge rating. Data and results from all phases of this project were presented at the American Geophysical Union meeting in Portland in October 2009 and at the Federal Interagency Sedimentation Conference in Las Vegas in June, 2010.