Resolving two‐dimensional flow structure in rivers using large‐scale particle image velocimetry: An example from a stream confluence

Abstract

Large-scale particle image velocimetry (LSPIV) has emerged as a valuable tool for measuring surface velocity in a variety of fluvial systems. LSPIV has typically been used in the field to obtain velocity or discharge measurements in relatively simple one-dimensional flow. Detailed two-dimensional or threedimensional characterization of flow structure has been relegated to laboratory settings because of the difficulty in controlling PIV limiting factors such as poor particle seeding, the need for camera rectification, and challenging field conditions. In this study we implement a low-cost LSPIV setup using a high-resolution action camera mounted above a stream confluence and water seeded with recycled landscape mulch. Time-averaged 2-D velocities derived from LSPIV are compared with those measured with an acoustic Doppler velocimeter (ADV) in the camera's field of view. We also assess the capabilities of this setup to resolve turbulent and time-averaged flow structures at a stream confluence. Our results reveal that even in challenging field conditions a basic LSPIV setup can yield accurate data on velocity and resolve in detail the temporal evolution of flow structures on the surface of rivers. The resulting dataset contains velocity information at high spatial and temporal resolution, a significant advance in understanding flow processes at stream confluences. Our LSPIV analysis provides support for previous numerical modeling studies that have distinguished between Kelvin-Helmholtz and wake modes of turbulent behavior within the mixing interface at confluences. This study shows that LSPIV can provide unprecedented levels of resolution of surface velocity patterns on rivers. Detailed velocity data derived from LSPIV can be used to evaluate numerical predictions of flow structure in complex fluvial environments.

Sand particle velocities over a subaqueous dune slope using high‐frequency image capturing

“…The DPIV-technique calculates particle displacement for particle groups. The analysis is sensitive to the choice of the size of the interrogation area (IA), among other pre-processing factors (Lewis and Rhoads, 2015). The analysis is sensitive to the choice of the size of the interrogation area (IA), among other pre-processing factors (Lewis and Rhoads, 2015).…”

“…Cross-correlations between particle groups in two subsequent images are evaluated and provide the most probable displacement of the particles in a subpart of the image, assuming the particles travel in a straight line from image A to image B (Raffel et al, 2007). Lewis and Rhoads (2015) show that a small IA and short time steps provide more details on particle displacement, but does not necessarily provide more accurate results. A different size of IA yields different velocity data time series for the same extraction location.…”

LSPIV Measurements of Two‐Dimensional Flow Structure in Streams Using Small Unmanned Aerial Systems: 2. Hydrodynamic Mapping at River Confluences

“…KRCS has an asymmetrical (y‐shaped) planform with a junction angle of 70°, whereas SALINE has an approximately symmetrical (Y‐shaped) planform with a junction angle of 70°. Past work at these sites has focused on measurements of flow structure using 2‐D or 3‐D velocity sensors at fixed cross sections, the relation of this flow structure to bed morphology and changes in this morphology, and on thermal mixing within the confluences (Lewis & Rhoads, ; Rhoads, ; Rhoads & Kenworthy, ; Rhoads & Kenworthy, ; Rhoads & Sukhodolov, , , ; Sukhodolov & Rhoads, ). KRCS has also served as the field site for several numerical modeling investigations of confluence hydrodynamics (Constantinescu et al, , , ; Constantinescu et al, ).…”

“…Although indirect evidence from field measurements of turbulence within the mixing interface at confluences is suggestive of the existence of these modes of vortex development (Rhoads & Sukhodolov, ; Sukhodolov & Rhoads, ), direct evidence is lacking and requires visualization of turbulent structures at confluences. Recent work suggests that LSPIV has the potential to provide direct evidence of different modes of vortex development within the confluence mixing interface (Lewis & Rhoads, , ).…”

A framework for modeling connections between hydraulics, water surface roughness, and surface reflectance in open channel flows

“…If this potential can be realized, passive optical remote sensing could complement other methods for noncontact flow measurement. Current approaches include large‐scale particle image velocimetry (PIV) using foam or other debris captured by video [e.g., Muste et al , ; Lewis and Rhoads , ; Tauro et al , ] or thermal signatures [e.g., Puleo et al , ; Dugan et al , ], various forms of radar [e.g., Costa et al , ; Fulton and Ostrowski , ], and airborne acoustic measurements of surface roughness [ Krynkin et al , ]. The ability to estimate velocities from optical, thermal, radar, or acoustic data could enable remote measurement of river discharge, provided depth information also can be obtained via remote sensing or some other independent means; this is an important and long‐standing objective in the hydrologic sciences.…”