Cumulative impacts of river engineering, Mississippi and Lower Missouri rivers

Abstract

The goal of this study was to construct a large, data-rich model to test hydrological responses to engineering modifications on over 3200 km of the Mississippi and Lower Missouri Rivers. We compiled model explanatory variables from a geospatial database quantifying construction of all bridges, wing dikes, bendway weirs, levees, artificial meander cutoffs, channel constriction and navigational dams over the past 100-150 years. Response variables were derived from 68 rated and un-rated hydrologic stations in the study area, with responses analysed across a range of discharges from within-channel flows up to moderate floods. Correlation analysis, multiple linear regression and stepwise regression analyses document strong and consistent responses to construction history, both in individual reach-scale models and systemwide. Meander cutoffs are associated with degradation and acceleration of flow that has reduced stages across the full discharge range. Navigational dams on the Upper Mississippi River increased low-flow stages and flood levels to a lesser extent, with little or no post-dam change. One of the strongest signals was the hydrologic response to wing-dike construction, which resulted in large back-water increases in stage upstream of wing dikes and mixed effects downstream, including the overlapping effects of incision and velocity losses. Levees were associated with local flow concentration, overbank storage loss and floodplain conveyance loss depending on reachscale conditions. The results presented here (1) quantify incremental and cumulative hydrologic responses to a range of engineering activities and (2) provide an empirical tool for verifying and assessing hydraulic and other models of river-system change.

Hydraulic and flood-loss modeling of levee, floodplain, and river management strategies, Middle Mississippi River, USA

“…Between 1945 and the present, the total length of wing dikes on the MMR increased from 120,400 m (cumulative length ;Maher 1964) to at least 344,000 m (Pinter et al 2010). Between 1990and 1998,500 m of new bendway weirs were built (Pinter et al 2010), and a wave of new ''chevron dikes'' are now being built and planned along the MMR (see Sect. 1).…”

“…Based on hydraulic principles, Yen (1995) found that for a given flood, levees raise the stage in the channel adjacent to a levee and upstream of it. On the MMR, Pinter et al (2008Pinter et al ( , 2010 and Remo et al (2009) found that levee construction was the second most important contributor to historical flood-level increases, after channel training structures. Dyhouse (1985) assessed the impact of agricultural levees on MMR flood stages, suggesting a 1.5-m levee-driven increase for the 100-year flood.…”

“…Stevens et al (1975) looked at the geomorphic characteristics of the MMR and found that changes in cross-sectional area due to navigation structures reduced the flow carrying capacity of the river. Pinter et al (2008Pinter et al ( , 2010 developed multivariate statistical models to identify the causal mechanisms driving observed changes in river stage over time. They concluded that the largest factors contributing to increased flooding on the MMR were wing dikes and related navigation structures (Pinter et al , 2010.…”

“…Pinter et al (2008Pinter et al ( , 2010 developed multivariate statistical models to identify the causal mechanisms driving observed changes in river stage over time. They concluded that the largest factors contributing to increased flooding on the MMR were wing dikes and related navigation structures (Pinter et al , 2010. Remo et al (2009) modeled the effects of potential contributors to increased flood frequencies-levee expansion, channel modifications, and land-cover changes-and concluded that 38-70% of changes in flood stage on the MMR were attributable to changes in channel geometry and hydraulic roughness driven by navigation construction in the channel.…”

“…The levee footprint and protection levels for the present data levees were obtained from the Scientific Assessment and Strategy Team's (SAST) levee (SAST 1995). Levee locations and elevations for Scenario 4 were from the database described by Remo et al (2008) and Pinter et al (2008Pinter et al ( , 2010. Data requirements for the Hazus-MH flood-loss model include topographic, infrastructure, and damage-depth curve.…”