Hostname: page-component-7b7567c8c-z4zkx Total loading time: 0 Render date: 2026-11-16T01:10:32.886Z Has data issue: false hasContentIssue false

Internal wave boluses as coherent structures in a continuously stratified fluid

Published online by Cambridge University Press:  06 January 2020

Guilherme S. Vieira
Affiliation:
Department of Mechanical and Industrial Engineering, Northeastern University, Boston, MA02115, USA
Michael R. Allshouse*
Affiliation:
Department of Mechanical and Industrial Engineering, Northeastern University, Boston, MA02115, USA
*
†Email address for correspondence: m.allshouse@northeastern.edu

Abstract

Internal waves shoaling on the continental slope can break and form materially coherent vortices called boluses. These boluses are able to trap and transport material up the continental slope, yet the global extent of bolus transport is unknown. Previous studies of bolus formation primarily focused on systems consisting of two layers of uniform density, which do not account for the presence of ocean pycnoclines of finite thickness. We use hyperbolic tangent profiles to model the density stratification in our simulations and demonstrate the impact of the pycnocline on the bolus. A spectral clustering method is used to objectively identify the bolus as a Lagrangian coherent structure that contains the material advected upslope. The bolus size and displacement upslope are examined as a function of the pycnocline thickness, incoming wave energy, density change across the pycnocline and topographic slope. The dependence of bolus transport on the pycnocline thickness demonstrates that boluses in continuous stratifications tend to be larger and transport material further than in corresponding two-layer stratifications.

Information

Type
JFM Papers
Copyright
© 2020 Cambridge University Press
Figure 0

Figure 1.Potential density profiles $\unicode[STIX]{x1D70C}_{\unicode[STIX]{x1D703}}(z)$ near the continental shelf obtained from the World Ocean Circulation Experiment Hydrographic Program (Schlitzer 2000). (a–d) Observational data (black dots) and the best fit line using a hyperbolic tangent function (coloured line). The pycnocline thickness $\unicode[STIX]{x1D6FF}$ describes the width of the transition in density. (e) The geolocation of each density profile is identified with a marker corresponding to the colour of the best fit line: (a) red, (b) green, (c) blue and (d) yellow.

Figure 1

Figure 2.Schematic diagram of the computational domain, dimensions, mesh resolution, density profile (in red) and fundamental mode forcing velocity profile (in blue). The distance from the inlet boundary on the left to the midpoint of the slope $L=2.87~\text{m}$ and the domain height $H=0.4~\text{m}$ are constant in all simulations, while $\unicode[STIX]{x1D6FF}$ and $s$ are varied in the parametric study. A sample background density profile $\unicode[STIX]{x1D70C}_{0}(z)$ (red) and inlet velocity profile $U(z)$ (blue) are presented corresponding to the sample simulation ($\unicode[STIX]{x1D6FF}=0.2~\text{m}$, $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}=20~\text{kg}~\text{m}^{-3}$, $s=0.176$, presented in § 2.2). The mesh resolution of the sample simulation is presented in a grey logarithmic scale with the highest resolution $\unicode[STIX]{x1D6E5}_{x}=2\times 10^{-4}~\text{m}$ in the breaking region (black) and the lowest resolution $3\times 10^{-3}~\text{m}$ in the evanescent region (light grey).

Figure 2

Figure 3.(a) Background density stratification for a density change $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}=20~\text{kg}~\text{m}^{-3}$ and pycnocline thicknesses $\unicode[STIX]{x1D6FF}=0.025$ (blue), 0.2 (red) and 0.4 m (green), as well as $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}=10~\text{kg}~\text{m}^{-3}$ and $\unicode[STIX]{x1D6FF}=0.2~\text{m}$ (yellow). (b) The normalized mode-1 velocity profile amplitude $U(z)$ for each corresponding stratification, with frequency $\unicode[STIX]{x1D714}=0.628~\text{rad}~\text{s}^{-1}$. The red horizontal dashed lines indicate the pycnocline thickness for $\unicode[STIX]{x1D6FF}=0.2~\text{m}$, and the yellow vertical dashed lines the density change for $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}=10~\text{kg}~\text{m}^{-3}$.

Figure 3

Figure 4.(a) Time evolution of the mid-depth isopycnal, $\unicode[STIX]{x1D70C}_{H/2}=1010~\text{kg}~\text{m}^{-3}$, with the wave amplitude scale bar indicated in red. The dashed lines represent the unperturbed isopycnal height. (b) Instantaneous density perturbation field $\unicode[STIX]{x1D70C}^{\prime }$ of the sample simulation ($\unicode[STIX]{x1D6FF}=0.2~\text{m}$) at $t=55~\text{s}$ for the full domain and (c) breaking region. Positive values (red) represent fluid displaced upward and negative values (blue) represent fluid displaced downward. The solid grey lines in (b) represent isopycnals. The isopycnal $\unicode[STIX]{x1D70C}=1005~\text{kg}~\text{m}^{-3}$ is drawn in black and used to highlight the bolus front boundary in (c,d). The breaking region presented in (c) corresponds to the region surrounded by the thick black box in (b). (d) Velocity field corresponding to the region surrounded by the thick black box in (c), which contains the bolus front. (Video of the evolution of (b,c) is available in the supplementary material.)

Figure 4

Figure 5.(a) Sample simulation clustering result for elements in the initial uniform grid. Seven clusters have been identified, with the Lagrangian bolus cluster represented in dark blue. (b) The time evolution for the bolus cluster from $t_{0}=19.25~\text{s}$ to $t_{f}=65~\text{s}$, which is the entire bolus lifespan. (Video of the evolution of the figure is available in the supplementary material.)

Figure 5

Figure 6.Kinematic comparison between the bolus and a co-propagating non-breaking isopycnal above the breaking region for $t\in [t_{0},t_{f}]$. (a) Bolus and amplified isopycnal for three time instances. The wave vertical positions and amplitudes are modified for illustration purposes. (b) The horizontal position as a function of time and (c) the horizontal speed as a function of horizontal position for the leading wave trough (dashed blue), the leading wave crest (dashed red), the bolus centre of volume (bold black), bolus trailing point (thin black) and foremost point (bold dark red). The isopycnal used to track the crest and trough in this plot was $\unicode[STIX]{x1D70C}=1000.05~\text{kg}~\text{m}^{-3}$, which has a neutrally buoyant height at $z=0.3635~\text{m}$. The tick marks in (b) correspond to the times for which the bolus and isopycnals are plotted in (a).

Figure 6

Figure 7.(a) Time evolution of the three-dimensional bolus. The tracer lateral coordinate $\widehat{y}$ at $t_{0}=19.25~\text{s}$ is indicated in colour (from red to blue), and the different levels of shading indicate different time instances. (b) Projected two-dimensional (2-D; red) and three-dimensional (3-D; grey) bolus tracers at the initial time $t_{0}$ and (c) at $t=55~\text{s}$. Transparency is used to indicate lower or higher concentration of tracers in the projected views.

Figure 7

Table 1.The bolus simulation cases considered in the respective sections and figures are presented here. The pycnocline thickness $\unicode[STIX]{x1D6FF}$, the energy at the breaking point $E_{k}$, the change in density $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}$ and the topographic slope $s$ are independently varied for each of the parameter studies. (An expanded version of this table is available in the supplementary material.)

Figure 8

Figure 8.Time evolution of the boluses from $t_{0}$ to $t_{f}$ for stratifications with pycnocline thicknesses (a) $\unicode[STIX]{x1D6FF}=0.025~\text{m}$, (b) 0.1 m, (c) 0.2 m, (d) 0.25 m and (e) 0.3 m. The trajectory of the bolus centre of volume for $\unicode[STIX]{x1D6FF}=0.2~\text{m}$ and the displacement upslope $D_{b}$ are illustrated in (c). (Video of the evolution of the figure is available in the supplementary material.)

Figure 9

Figure 9.(a) Bolus size $S_{b}$ and (b) maximum displacement upslope $D_{b}$ as a function of the pycnocline thickness, $\unicode[STIX]{x1D6FF}$, for mode-1 waves breaking with constant energy. The error bars represent the range of displacement within the bolus tracers. Different markers are used to represent bolus shape categories: ball (circle), hook (triangle) and sliver (square).

Figure 10

Figure 10.(a) Ball, (b) hook and (c) sliver bolus shape categories represented by the bolus tracer position at time $t_{0}$. Shapes are obtained for $\unicode[STIX]{x1D6FF}=0.025~\text{m}$ (blue), 0.25 m (green) and 0.3 m (red). The aspect ratio is approximately 30:1. The markers in the top right of each plot are used to represent bolus shape categories.

Figure 11

Figure 11.(a) Bolus size $S_{b}$ and (b) maximum displacement upslope $D_{b}$ as a function of the energy factor for the shoaling wave. Simulations were performed with the pycnocline thicknesses $\unicode[STIX]{x1D6FF}=0.025$ (blue), 0.2 (black) and 0.4 m (red). When no bolus is identified, a cross is plotted at $S_{b}=0~\text{m}^{2}$ and $D_{b}=0~\text{m}$. Marker shapes are used to represent each bolus shape categories: ball (circle), hook (triangle) and sliver (square).

Figure 12

Figure 12.(a) Bolus size $S_{b}$ and (b) maximum displacement upslope $D_{b}$ as a function of the pycnocline thickness, for variable density change, $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}$. Simulations were performed with density changes $\unicode[STIX]{x0394}\unicode[STIX]{x1D70C}=10$ (red), 20 (black), 40 (blue) and $80~\text{kg}~\text{m}^{-3}$ (green). When no bolus is identified, a cross is plotted at $S_{b}=0~\text{m}^{2}$ and $D_{b}=0~\text{m}$. Marker shapes are used to represent each bolus shape categories: ball (circle), hook (triangle) and sliver (square).

Figure 13

Figure 13.(a) Bolus size $S_{b}$, (b) maximum displacement upslope $D_{b}$ and (c) maximum vertical displacement $\unicode[STIX]{x0394}z_{b}$ as a function of the topographic slope $s$ for the pycnocline thicknesses $\unicode[STIX]{x1D6FF}=0.025$ (blue), 0.2 (black), 0.4 m (red). (d) Bolus size $S_{b}$, (e) maximum displacement upslope $D_{b}$ and (f) maximum vertical displacement $\unicode[STIX]{x0394}z_{b}$ as a function of the pycnocline thickness $\unicode[STIX]{x1D6FF}$, for slopes $s=0.105$ (red), 0.176 (black), 0.231 (blue). Marker shapes are used to represent each bolus shape category: ball (circle), hook (triangle) and sliver (square). Empty red markers for $s=0.105$, $\unicode[STIX]{x1D6FF}=0.3~\text{m}$ represent two possible values for the bolus at the hook-sliver transition.

Figure 14

Figure 14.Relationship between dimensionless parameters and bolus size $S_{b}$ and distance upslope $D_{b}$. Marker sizes represent the magnitude of the plotted quantities, and colours represent different studies: pycnocline thickness variation (grey), energy variation (green), density change variation (blue), slope variation (red). Crosses represent cases for which no bolus is formed.

Viera et al. supplementary movie 1

Density perturbation field ρ' of the sample simulation presented in section 2.2 for the (top) full domain and (bottom) breaking region. Positive values (red) represent fluid displaced upward and negative values (blue) represent fluid displaced downward. The solid gray lines represent isopycnals. The isopycnal ρ=1015 kg/m3 is drawn in black and used to highlight the bolust front boundary. The breaking region (bottom) corresponds to the region surrounded by the thick black box in the full domain (top).

Download Viera et al. supplementary movie 1(Video)
Video 26 MB

Viera et al. supplementary movie 2

Density perturbation field ρ' of the sample simulation presented in section 2.2 for the (top) full domain and (bottom) breaking region with passive tracers distributed outside and inside the breaking zone. The tracers in the constant depth region temporarily oscillate vertically as the wave passes. Tracers in the breaking region are entrained in the resulting vortex demonstrating how the breaking mechanism results in effective transport.

Download Viera et al. supplementary movie 2(Video)
Video 37 MB

Viera et al. supplementary movie 3

(top) Evolution of the sample simulation tracers with the tracer color determined by cluster membership. Seven clusters have been identified, with the Lagrangian bolus cluster represented in green. (bottom) The time evolution for the bolus cluster from t0 = 19.25s to tf = 65s. Instantaneous positions of the bolus presented in figure 5(b) are presented here as well. The trajectory of the bolus center of volume is illustrated in gray.

Download Viera et al. supplementary movie 3(Video)
Video 30.8 MB

Viera et al. supplementary movie 4

Time evolution of the boluses, from t0 to tf indicated by the time values on the left and right, for stratifications with pycnocline thicknesses δ = 0.025, 0.1, 0.2, 0.25 and 0.3m. This movie corresponds to a dynamic view of the data presented in figure 8.

Download Viera et al. supplementary movie 4(Video)
Video 45.5 MB
Supplementary material: PDF

Vieira et al. supplementary material

Supplementary table

Download Vieira et al. supplementary material(PDF)
PDF 109.7 KB
Supplementary material: PDF

Vieira et al. supplementary material

Supplementary material A.

Download Vieira et al. supplementary material(PDF)
PDF 146.6 KB
Supplementary material: PDF

Vieira et al. supplementary material

Supplementary material B.

Download Vieira et al. supplementary material(PDF)
PDF 143.8 KB
Supplementary material: PDF

Vieira et al. supplementary material

Supplementary material C.

Download Vieira et al. supplementary material(PDF)
PDF 203 KB
Supplementary material: PDF

Vieira et al. supplementary material

Supplementary material D.

Download Vieira et al. supplementary material(PDF)
PDF 110.7 KB
Supplementary material: PDF

Vieira et al. supplementary material

Additional movie descriptions

Download Vieira et al. supplementary material(PDF)
PDF 124.2 KB