Effects of streamflows on stream-channel morphology in the eastern Niobrara National Scenic River, Nebraska, 1988–2010

The Niobrara River is an important and valuable economic and ecological resource in northern Nebraska that supports ecotourism, recreational boating, wildlife, fisheries, agriculture, and hydroelectric power. Because of its uniquely rich resources, a 122-kilometer reach of the Niobrara River was designated as a National Scenic River in 1991, which has been jointly managed by the U.S. Fish and Wildlife Service and National Park Service. To assess how the remarkable qualities of the National Scenic River may change if consumptive uses of water are increased above current levels, the U.S. Geological Survey, in cooperation with the National Park Service, initiated an investigation of how stream-channel morphology might be affected by potential decreases in summer streamflows. The study included a 65-kilometer segment in the wide, braided eastern stretch of the Niobrara National Scenic River that provides important nesting habitat for migratory bird species of concern to the Nation.

The study focused on three river segments, separated at the confluences with two tributaries, Plum Creek and Long Pine Creek. With an overall temporal scope of 1988–2010 that includes a short interval preceding and a long interval following the Niobrara National Scenic River Designation Act of 1991, the study analyzed five separate time periods: 1988–93, 1994–99, 2000–3, 2004–6, and 2007–10, each of which ended with a year in which aerial photography coverage was available.

Streamflow duration was analyzed for one streamgage upstream from the study area and two streamgages on tributary streams within the study area. Summer streamflows (July, August, and September) were targeted for analysis because median flows of the Niobrara River are lowest during those 3 months. In addition, peak flows during the study period were used to estimate bankfull discharge, which is one determinant of channel dimensions.

Changes in channel morphology were examined using aerial photographs from 1993, 1999, 2003, 2006, and 2010 to measure channel width, area of islands, and incipient flood-plain surfaces, and to compute the braided index. Channel metrics were computed for each photography year and summarized by river segment. Additionally, at fixed-location cross sections, photography analysis identified localized geomorphic change to infer processes. Accuracy of geomorphic feature classification was estimated and the root-mean-square difference (RMSD) between aerial photographs was calculated to determine associated errors in channel metric calculations. The horizontal accuracy of boundaries delineated in the classification was estimated as 5 meters (m) for boundaries based on 1993 aerial photography and 4 m for all other aerial photography. The RMSD between aerial photography years ranged from 3.04 m to 4.16 m.

The largest measurable changes in channel metrics were measured between 1993 and 1999 and between 1999 and 2003. Between 1993 and 1999, average total channel width increased by 9 m (3 percent) and 14 m (5 percent) in segments 2 and 3, respectively; average active channel width increased by 13 m (5 percent) in segment 3 and decreased by 6 m (4 percent) in segment 1; and incipient flood-plain-surface area increased by 40, 44, and 33 percent in segments 1, 2, and 3, respectively. Changes in channel metrics between 1999 and 2003 included a decrease in average total channel width of 14 m (5 percent) in segment 2; a decrease in active channel widths of 8 m (3 percent) and 6 m (2 percent) in segments 2 and 3, respectively; and an increase of 5 m (3 percent) in segment 1. Incipient flood-plain areas decreased by 22 and 33 percent in segments 1 and 2, respectively, and increased by 42 percent in segment 3.

Large changes were measured between 1993 and 1999, and between 1999 and 2003, at many of the fixed-location cross sections. Large changes (that is, greater than 25 percent) in total channel width were measured in all three segments between 1993 and 1999 and again between 1999 and 2003; large increases were dominant between 1993 and 1999 and large decreases were dominant between 1999 and 2003. Segment 1 was the most susceptible to localized changes as there was only one period (between 2003 and 2006) in which the active channel width largely changed in fewer than 10 percent of the cross sections.

Changes in channel metrics generally corresponded to changes in streamflow conditions, but other than changes in incipient flood-plain area, these changes were small and were not measured in all three segments simultaneously. Increases in total channel width (except in segment 1) and incipient flood-plain area between 1993 and 1999 corresponded to increases in streamflow. Channel narrowing (except in segment 1) between 1999 and 2003 corresponded to lower summer streamflows and extended durations of very low summer streamflow. Although the pattern of low summer streamflow and extended durations of very low summer streamflow continued during the 2004–6 period and at the beginning of the 2007–10 period, no further narrowing was measured. Consistent tributary summer inflows help to explain the resistance of segments 2 and 3 to further narrowing. Because segment 1 is already much narrower than segments 2 and 3, its average current velocity is likely to be swifter and, therefore, competent to offset further effects of the processes that led to its narrowness.