Spatial Patterns of Channel Instability Along an Ozark River, Southwest Missouri

Abstract

In the Ozark Highlands of Missouri, unstable river reaches that display rapid planform change are described as active reaches. While active reaches can be part of the natural morphodynamic regime, accelerated gravel bar deposition and bank erosion have been linked to historical and recent anthropogenic activities. Relationships between geomorphic controls and specific forms of channel instability are poorly understood in the Ozarks. The objectives of this research were to (1) develop an active reach classification scheme that provides an objective means of identifying and evaluating longitudinal patterns of instability in this and other Ozark rivers, (2) identify active reaches and different forms of instability along 80 km of the Finley River in southwest Missouri, and (3) investigate physical controls on active reaches. Historical aerial photographs and geographic information system (GIS) analysis showed that active reaches occur along 21% of the length of the main stem, preferentially in valleys with confinement ratios between 10 and 30 and near major tributary confluences. Four active reach forms were identified with proportion of length as follows: extension (8%), megabar (6%), cutoff (5%), and translation (2%). Depositional megabar-type active reaches tended to form directly downstream of erosional extensiontype active reaches, probably due to excess gravel supply by reworking of historical floodplain deposits. The lack of a dominant active reach form, along with a stable main stem sinuosity over time, suggests that the Finley River is a semi-stable, self-organized system in balance with watershed inputs since at least 1955, the date of the earliest photo series.

Anthropogenic landforms and sediments from dredging and disposing sand along the Apalachicola River and its floodplain

“…Floodplain sand and gravel mining activity, particularly when not well regulated, has increased bar size due to avulsions, increased erosion and flooding over bare landscapes (Mossa and Marks, 2011). Rapid channel migration can form large, unvegetated gravel bars, in part related to the legacy of historical human activities (Martin and Pavlowsky, 2011). In the Kissimmee River, sand bars grew once the canal C-38 was backfilled and flow was reactivated in the original channel; this was due to water inputs and associated processes of erosion and deposition in former stagnant areas (Mossa, 2015).…”

Investigating patterns of in-channel wood deposition locations in a low-gradient, variably-confined alluvial river system

“…Gravel bar area was determined from the digitized gravel bars. Although gravel bar area is dependent on stage at the time of the photograph, and is very dynamic in many fluvial settings, using this metric can be justified in the following ways: 1) Previous research by the authors (Martin and Pavlowsky, 2011) indicates that morphologically similar Ozark streams are equilibrated to their current sediment inputs resulting in little to no change in bar locations over decadal time scales; 2) differences in stage height (þ 0.3 m), as measured at the United States Geological Survey gage station at Irondale, Missouri (Gage number 07017200), between data collection dates and aerial photograph acquisition dates were deemed negligible because the resulting difference in exposed gravel bar area was likely within the margin of aerial photograph user digitizing error. Table 1 provides a description of each of these variables.…”

“…The Big River exemplifies a low-energy, variably-confined meandering river system and is morphologically representative of typical Ozark-region river systems which display similar channel confinement characteristics (Martin and Pavlowsky, 2011; Martin et al, 2016; Owen et al, 2011; Pavlowsky et al, 2017). The Big River is considered variably-confined due to the irregular variation of valley widths (and thus channel confinement) in the downstream direction (Figure 2).…”

Measuring channel planform change from image time series: A generalizable, spatially distributed, probabilistic method for quantifying uncertainty

“…A more complex approach to quantifying uncertainty is to establish a level‐of‐detection (LoD); measurements of channel change that are smaller in magnitude than this threshold cannot be distinguished from uncertainty and are removed from the analysis (Urban and Rhoads, 2003). In most studies, the LoD is specified as a spatially uniform threshold for designating measurements as non‐significant and excludes these measurements from the analysis (Winterbottom and Gilvear, 2000; White et al ., 2010; Martin and Pavlowsky, 2011; Kessler et al ., 2013). This approach causes a large number of small planform changes to be removed from the analysis and introduces a bias by ignoring polygons of very small channel change, implying that the reach‐scale average will be dominated by polygons of larger channel change.…”