Search

Since the early 20th century, the Green River, the longest tributary of the Colorado River, has narrowed, decreasing available riparian and aquatic habitat. Initially, the widespread establishment of non-native tamarisk was considered to be the primary driver of channel narrowing. An alternative hypothesis postulated that changes in hydrology drove narrowing. Reductions in total streamflow and changes to flow regime occurred due to wide-spread water development, decreased snowmelt flood magnitude, and the increased cyclicity of wet and dry years. The two hypotheses agree on channel narrowing, but each influences modern river management differently. A tamarisk-driven model of narrowing implies that modern flow management doesn’t substantially affect channel change. Conversely, channel narrowing driven by changes in hydrology implies that present flow management decisions matter and continued adjustments to flow regime may result in future channel change. To understand the roles of decreasing total annual flow, declining annual peak flood magnitude, and changing vegetation communities on 20th century channel narrowing, we investigated channel narrowing along the lower Green River within Canyonlands National Park (CNP). Previous studies agree that the channel has narrowed, however, the rate, timing and magnitude of documented narrowing are only partially understood. Multiple lines of evidence were used to reconstruct the history of channel narrowing in the lower Green River. This study focuses on channel narrowing, but additionally investigated possible changes to channel depth, identified process, timing and magnitude of floodplain formation. Floodplain formation was described in the field using stratigraphy, sedimentology, and dendrogeomorphology exposed in a floodplain trench. Channel and floodplain surveys were conducted to determine possible changes in bed elevation. Additionally, existing aerial imagery, hydrologic data, and sediment transport data were analyzed. These techniques were applied to determine magnitude, timing and processes of channel narrowing at multiple spatial and temporal scales. The floodplain investigation identified a new period of channel narrowing by vertical accretion after high peak flow years of 1983 and 1984. Narrowing was initiated by vertical accretion in the active channel, deposited by moderate floods exceeded more than 50% of the time. Vertical accretion continued in the early 1990s, converting the active channel into a periodically inundated floodplain surface. Suspended-sediment deposition dominated deposits, resulting in the formation of natural levees and floodplain troughs in both inset floodplains. Rates of deposition were highly variable, ranging from 0.03-0.50 m/yr. The lower Green River within Canyonlands National Park has narrowed substantially since the late 1800s, resulting in a narrower channel. Changes to flood magnitude, rate and timing since 1900, driven by increased water storage and diversion in the Green River basin and declines in annual precipitation, were responsible for inset floodplain formation. Floodplains of the contemporary lower Green River in CNP began forming in the late 1930s and continued to form and vertically aggrade in the 20th century by inset floodplain formation. During this time period, peak flow and total runoff declined due to climatic changes and water development. Analysis of aerial imagery covering 61 kilometers (km) of the Green River in CNP shows that changes to the floodplain identified in the trench are representative of the entire study area. The establishment of non-native tamarisk (Tamarix spp.) did not drive channel narrowing, though dense stands stabilized banks and likely promoted sediment deposition. The lower Green River narrowed 12% from 1940-2014, with the majority of narrowing (10% of all narrowing) occurring from the 1980s to the present. Inset floodplain formation reflects changes to flood magnitude and timing resulting from water development and decreases in natural runoff. Findings suggest that long-term management of the riverine corridor within Canyonlands National Park will require a greater focus on upstream flow contributions and how those flows are currently managed. Recovery of endangered endemic native fishes, the Colorado pikeminnow (Ptychocheilus lucius), and the razorback sucker (Xyrauchen texanus), plays a primary role in determining current flow allocations. Collaboration with upstream stakeholders and managers is necessary to maximize elements of the flow regime that preserve channel width and limit channel narrowing.

Documentation of the interacting effects of river regulation and climate on riparian vegetation has typically been limited to small segments of rivers or focused on individual plant species. We examine spatiotemporal variability in riparian vegetation for the Colorado River in Grand Canyon relative to river regulation and climate, over the five decades since completion of the upstream Glen Canyon Dam in 1963. Long-term changes along this highly modified, large segment of the river provide insights for management of similar riparian ecosystems around the world. We analyze vegetation extent based on maps and imagery from eight dates between 1965 and 2009, coupled with the instantaneous hydrograph for the entire period. Analysis confirms a net increase in vegetated area since completion of the dam. Magnitude and timing of such vegetation changes are river stage-dependent. Vegetation expansion is coincident with inundation frequency changes and is unlikely to occur for time periods when inundation frequency exceeds approximately 5%. Vegetation expansion at lower zones of the riparian area is greater during the periods with lower peak and higher base flows, while vegetation at higher zones couples with precipitation patterns and decreases during drought. Short pulses of high flow, such as the controlled floods of the Colorado River in 1996, 2004, and 2008, do not keep vegetation from expanding onto bare sand habitat. Management intended to promote resilience of riparian vegetation must contend with communities that are sensitive to the interacting effects of altered flood regimes and water availability from river and precipitation. å©2015. American Geophysical Union. All Rights Reserved.

Longitudinal profile , channel cross-section geometry, and depositional patterns of the Green River in its course through the eastern Uinta Mountains are each strongly influenced by river-level geology and tributary sediment delivery processes. We surveyed channel cross sections at 1-km intervals, mapped surficial geology , and measured size and characteristics of bed material in order to evaluate the geomorphic organization of the 70- km study reach . Canyon reaches that are of high gradient and narrow channel geometry are associated with the most resistant lithologies exposed at river level and the most frequent occurrences of tributary debris fans. Meandering reaches that are characterized by low gradient and wide channel geometry are associated with river-level lithology that is of moderate to low resistance and very low debris fan frequency. The channel is in contact with bedrock or talus along only 42 percent of the bank length in canyon reaches and there is an alluvial fill of at least 12 m that separates the channe l bed from bedrock at three borehole sites. The influence of lithology primarily operates through the presence of resistant boulders in debris fans that are delivered by debris flows from steep tributaries. The depositional settings created by debris fans consist of (1) channel-margin deposits in the backwater above the debris fan, (2) eddy bars in the zone of recirculating flow below the constriction, and (3) expansion gravel bars in the expansion below the zone of recirculating flow. These fan-eddy complexes are the storage location of about 70 percent, by area, of all fine- and coarse-grained alluvium contained within the canyons above the low-water stage. Immediately adjacent meandering reaches contain an order of magnitude more alluvium by area but have no debris fan-created depositional settings. This study also describes the flood-plain and terrace stratigraphy of the Green River in the eastern Uinta Mountains and changes due to the operations of Aarning Gorge Dam, upstream from the study area. These landforms are vertically aggrading deposits that are longiuidinally correlative throughout the 65-km study reach. The suite of surfaces identified includes a terrace that is inundated by rare pre- or post-dam floods, an intermediate bench that is inundated by rare post-dam floods, and a post-dam flood plain that is inundated by the post-dam mean annual flood. Analysis of historical photographs in the study reach shows that both the intermediate bench and post-dam flood plain are landforms that were not present in any of the 6 years for which photographs were examined between 1871 and 1954. Photographic replications also show that gravel bars consisting of bare gravel in 1922 and earlier photographs are now covered by fine-grained alluvium and vegetation. Decreased gravel-bar mobility is indicated by estimates of critical and average boundary shear stress. Comprehensive surficial geologic mapping of the study area indicates that the bankfull channel has decreased in width by an average of about 20 percent.