Abstract

The Merced River in the popular and picturesque eastern-most part of Yosemite Valley in Yosemite National Park, California, USA, has been extensively altered since the park was first conceived in 1864. Historical human trampling of streambanks has been suggested as the cause of substantial increases in stream width, and the construction of undersized stone bridges in the 1920s has been suggested as the major factor leading to an increase in overbank flooding due to deposition of bars and islands between the bridges. In response, the National Park Service at Yosemite National Park (YNP) requested a study of the hydraulic and geomorphic conditions affecting the most-heavily influenced part of the river, a 2.4-km reach in eastern Yosemite Valley extending from above the Tenaya Creek and Merced River confluence to below Housekeeping Bridge. As part of the study, present-day conditions were compared to historical conditions and several possible planning scenarios were investigated, including the removal of an elevated road berm and the removal of three undersized historic stone bridges identified by YNP as potential problems: Sugar Pine, Ahwahnee and Stoneman Bridges. This Open-File Report will be superseded at a later date by a Scientific Investigations Report.

A two-dimensional hydrodynamic model, the USGS FaSTMECH (Flow and Sediment Transport with Morphological Evolution of Channels) model, within the USGS International River Interface Cooperative (iRIC) model framework, was used to compare the scenarios over a range of discharges with annual exceedance probabilities of 50-, 20-, 10-, and 5- percent. A variety of topographic and hydraulic data sources were used to create the input conditions to the hydrodynamic model, including aerial LiDAR (Light Detection And Ranging), ground-based LiDAR, total station survey data, and grain size data from pebble counts. A digitized version of a historical topographic map created by the USGS in 1919, combined with estimates of grain size, was used to simulate historical conditions, and the planning scenarios were developed by altering the present-day topography. Roughness was estimated independently of measured water-surface elevations by using the mapped grain-size data and the Keulegan relation of grain size to drag coefficient. The FaSTMECH hydrodynamic model was evaluated against measured water levels by using a 130.9 m³ s^-1 flow (approximately a 33-percent annual exceedance probability flood) with 36 water-surface elevations measured by YNP personnel on June 8, 2010. This evaluation run had a root mean square error of 0.21 m between the simulated- and observed water-surface elevations (less than 10 percent of depth), though the observed water-surface elevations had relatively high variation due to the strong diurnal stage changes over the course of the 4.4-hour collection period, during which discharge varied by about 15 percent. There are presently no velocity data with which to test the model. A geomorphic assessment was performed that consisted of an estimate of the magnitude and frequency of bedload and suspended-sediment transport at “Tenaya Bar”, an important gravel-cobble bar located near the upstream end of the study site that determines the amount of flow across the floodplain at the Sugar Pine – Ahwahnee bend. An analysis of select repeat cross-sections collected by YNP since the late 1980s was done to investigate changes in channel cross-sectional area near the Tenaya Bar site.

The results of the FaSTMECH models indicate that the maximum velocities in the present-day channel within the study reach are associated with Stoneman and Sugar Pine Bridges, at close to 3.0 m s^-1 for the 5-percent annual exceedance probability flood. The modeled maximum velocities at Ahwahnee Bridge are comparatively low, at between 1.5 and 2.0 m s^-1, most likely due to the bridge's orientation parallel to down-valley floodplain flows. The results of the FaSTMECH models for the bridge removal scenarios indicate a reduction in average velocity at the bridge sites for the range of flows by approximately 23-38 percent (Sugar Pine Bridge), 32-42 percent (Ahwahnee Bridge), and 33-39 percent (Stoneman Bridge), though a side channel of concern to YNP management did not appear to be substantially affected by the removal scenarios. In comparison to the historical data, the FaSTMECH results suggest that flows for present-day conditions do not inundate the floodplain until between the 50- and 20-percent annual exceedance probability flood, whereas historically, a large portion of the floodplain was inundated during the 50-percent annual exceedance probability flood. Modeled maximum velocities in the present-day channel commonly exceed 2.0 m s^-1, whereas with the historical scenario, modeled maximum in-channel velocities rarely exceeded 2.0 m s^-1. The geomorphic analysis of the magnitude-frequency of bedload and suspended-sediment transport suggests that at the important Tenaya Bar site, the majority of bed sediment is mobile during most snowmelt-dominated floods. In contrast to sediment transport capacity, the analysis of repeat cross-sections suggests that bedload sediment supply into the eastern Yosemite Valley may be quite different between rain-on-snow floods and snowmelt-dominated floods, potentially with most sediment supply occurring during rain-on-snow floods, such as the 1997 flood. In contrast, the magnitude-frequency analysis of bedload and suspended-sediment transport suggests that long-term bedload sediment transport is likely dominated by snowmelt floods, and suspended-sediment transport is relatively low compared to bedload transport.

Obtaining measured velocity data throughout the study reach would aid in model calibration, and thus would improve confidence in model results. Improved confidence in the model velocity results would allow additional substantial analyses of reach-scale effects of the planning scenarios and would enable the development of geomorphic models to evaluate the long-term geomorphic responses of the site. In addition, the collection of watershed sediment-supply data, about which little is presently known, would give planners helpful tools to plan restoration scenarios for this nationally important river.

First posted January 14, 2013