Empirical Lagrangian parametrization for wind-driven mixing of buoyant particles at the ocean surface

Abstract

Abstract. Turbulent mixing is a vital component of vertical particulate transport, but ocean global circulation models (OGCMs) generally have low resolution representations of near-surface mixing. Furthermore, turbulence data is often not provided in reanalysis products. We present 1D parametrizations of wind-driven turbulent mixing in the ocean surface mixed layer, which are designed to be easily included in 3D Lagrangian model experiments. Stochastic transport is computed by Markov-0 or Markov-1 models, and we discuss the advantages/disadvantages of two vertical profiles for the vertical diffusion coefficient Kz. All vertical diffusion profiles and stochastic transport models lead to stable concentration profiles for buoyant particles, which for particles with rise velocities of 0.03 and 0.003 m s−1 agree relatively well with concentration profiles from field measurements of microplastics. Markov-0 models provide good model performance for integration timesteps of Δt ≈ 30 seconds, and can be readily applied in studying the behaviour of buoyant particulates in the ocean. Markov-1 models do not consistently improve model performance relative to Markov-0 models, and require an additional parameter that is poorly constrained.

Modelling submerged biofouled microplastics and their vertical trajectories

“…Our boundary condition at the surface is set such that if a particle is about to cross the surface, we set its depth to the NEMO-MEDUSA surface depth (0.6 m). Onink et al (2022) show that for positively buoyant particles, 1D vertical profiles estimated with this Markov-0 approach match reason- ably well with observations, with increased wind stress results in particles being mixed to greater depths and reduced particle rise velocities.…”

“…Wind-driven turbulence can play an important role in the vertical concentration profiles of buoyant particles (Kukulka et al, 2012), and therefore its inclusion is one of the novelties of this study (Table 1). Since we do not have access to the diffusivity profiles from NEMO-MEDUSA, we follow the approach from Onink et al (2022) to model turbulent stochastic transport in the surface mixed layer using a Markov-0 random walk model. The amount of turbulence in the surface mixed layer is computed using the K-profile parametrization (KPP) (Large et al, 1994;Boufadel et al, 2020), K z , given by…”

Size Dependent Transport of Floating Plastics Modeled in the Global Ocean

“…We intend to improve our model as research on the transport of ocean plastic continues. For example, an improved methodology for incorporating wind-driven mixing based on (Kukulka et al, 2012) into a Lagrangian particle tracking model such as ours is currently under review (Onink et al, 2021b); we plan to update our methodology to reflect theirs, as it has stronger physical motivation. Also, the work of (Ruiz et al, 2004) suggests wind-driven mixing could paradoxically serve to increase the concentration of micro (<1 mm) debris at the surface, since turbulent conditions can increase the rising velocity of these small particles.…”

“…Therefore, we would like to incorporate more coastal dynamics and model the effect of beaching in particular. Though recent research has made strong headway towards a preliminary understanding of the nearshore behavior of plastic debris (L. Lebreton et al, 2019;Olivelli et al, 2020;Morales-Caselles et al, 2021;Onink et al, 2021b;Ryan & Perold, 2021), these processes are still not well observed on a global scale. Coastal processes are complex, and more observation is essential to assess what mechanisms most strongly govern the nearshore behavior of plastic debris, and thus what mechanisms should be incorporated into global dispersal models such as ADVECT.…”

Bayesian comparison of stochastic models of dispersion

“…Both the Brownian and Langevin model can be extended to account for spatial anisotropy, inhomogeneity and the presence of a mean flow, at the cost of increasing the dimension of their parameter spaces; full details are given in Berloff & McWilliams (2002). Brownian and Langevin dynamics underlie the so-called random displacement and random flight models used for dispersion in the atmospheric boundary layer (Esler & Ramli 2017), and have been applied to the simulation of ocean transport, as models of mixing in the horizontal (Berloff & McWilliams 2002), vertical (Onink, van Sebille & Laufkötter 2022) and on neutral surfaces (Reijnders, Deleersnijder & van Sebille 2022). Ying, Maddison & Vanneste (2019) showed how Bayesian parameter inference can be applied to the Brownian model in the inhomogeneous setting using Lagrangian trajectory data.…”