Although the dynamics of secondary flow at river confluences have received considerable attention, a lack of theoretical insight in exploratory field and experimental research has led to different conclusions about the mechanisms driving these dynamics. This study revisits the problem of secondary flow at confluences by examining responses of the flow to controllable variations in curvature of flow trajectories in a series of field‐based experiments. The experiments were performed in a physical model designed to eliminate the influence of bed morphology, while using a large width‐to‐depth ratio to enable proper identification of flow structures at different scales. The results show that no secondary flow is detectable in runs with parallel merging flows irrespective of the momentum ratio. In runs with converging channels the secondary flow is represented by large‐scale motions consisting either of a pair of counterrotating helixes or a single helix depending on the momentum ratio. Comparison of scaled measured and theoretically predicted vertical profiles of lateral velocity shows that the helical secondary flow is primarily driven by the curvature of flow trajectories. This study also indicates that small‐scale secondary motions that have the same sense of rotation as the large‐scale helixes may develop in the immediate margins of the mixing interface. Differing only slightly from the flow depth in size, they are manifested by local upwelling within the large‐scale secondary motions. Several reasons to believe that interactions of large‐scale motions with opposing sense of rotation drive these small‐scale secondary flow cells are discussed in the paper.

No abstract

Only a handful of field studies have examined turbulent flow structure at discordant confluences; the dynamics of flow at such confluences have mainly been examined in the laboratory. This paper reports results of a field‐based investigation of turbulent flow structure at a discordant river confluence. These results support the hypothesis that flow at a discordant alluvial confluence with a velocity ratio greater than 2 exhibits jet‐like characteristics. Scaling analysis shows that the dynamics of the jet core are quite similar to those of free jets but that the complex structure of flow at the confluence imposes strong effects that can locally suppress or enhance the spreading rate of the jet. This jet‐like behavior of the flow has important implications for morphodynamic processes at these types of confluences. The highly energetic core of the jet at this discordant confluence is displaced away from the riverbed, thereby inhibiting scour; however, helical motion develops adjacent to the jet, particularly at high flows, which may promote scour. Numerical experiments demonstrate that the presence or absence of a depositional wedge at the mouth of the tributary can strongly influence detachment of the jet from the bed and the angle of the jet within the confluence.

Summary 1. Although macrophytes play a key role in the structure and functioning of lowland rivers, most of the basic plant, hydrodynamic and sediment‐water interactions have only been described qualitatively. We therefore studied quantitatively, the seasonal dynamics of matter deposition and mobilisation inside and outside (free path) a representative patch of arrowhead, Sagittaria sagittifolia, in the lowland River Spree, NE Germany, in August 2006. Our in situ study combined resuspension experiments, a hydrodynamically calibrated erosion chamber and concurrent measurements of the prevailing flow characteristics and bed load. 2. Increasing entrainment rates (E) of particles (ESPM) and total P (ETP), with increments of shear velocity (U*) from 0.53 to 2.42 cm s−1, were significantly higher inside the plant patch than outside. Indeed, ESPM and ETP at the lowest U* were 8‐ and 12‐fold higher inside than outside the patch, reflecting the resuspension potential of the upper nutrient‐enriched layer and the extent of pulsed P inputs even at small increases in U*. 3. Vertical distribution of velocity (u) revealed a flow pattern of a mixing layer inside the S. sagittifolia patch, and that of a boundary layer in the free path. The highest gradient of u in the mixing layer was located in the water column at about 0.5 m depth, whereas the highest gradient of u for the boundary layer was found near the riverbed. The maximum of U* (1.65 cm s−1) was only 4 mm above the sediment. 4. A plant mosaic provides a low‐energetic environment promoting extensive particle trapping and the accumulation of a fine‐grained, nutrient‐enriched sediment, and forming a large resuspension potential. Consequently, after plant decay and the concomitant increase of U* this material is preferentially entrained at higher rates. Hence, the key role of submerged macrophytes in lowland rivers is more directly related to modifying the dynamic equilibria between vegetation trapping and resuspension, than to the retention of nutrients, particularly P, and the reduction of P loads downstream to other waters.

[1] Dynamics of a vegetated mixing layer around a patch of submerged plants was examined theoretically and experimentally in a lowland river. Theoretical analysis explored analogy with hydrodynamic mixing layers and introduced equations that describe expansion of the layers and alterations in mean flow. Field experiments provided empirical data for validation of the theory and examination of the effects of population density on the mixing layer dynamics. Detailed measurements assessed the structure of turbulent flow, plant morphology, biomechanical properties, bending, and streamlining due to interaction with the flow. The theory was compared with the field experiments and an agreement was concluded. It was found that near the upstream edge of the patch, the vegetated mixing layer expands similar to the canonical mixing layers, although its statistics scale on the velocity differential characteristic for the stabilized part. Downstream mean momentum flux reduces in magnitude and reverses in direction, while turbulence became dominant. This was observed to stabilize the velocity differential and increase the mean velocity of the layer that result in its limited growth described by a logarithmic function. The population density of vegetation was found to control the flow. In dense vegetation, the analogy with mixing layers resembled best, while in sparse vegetation, the flow behaved similar to the boundary layers. This theory can be used for quantitative assessment of seasonal effects in natural vegetated rivers because the population density of vegetation varies during the vegetation season. The companion paper focuses on the structural properties of turbulence in the vegetated shear layers.