Search for: All records

Abstract Diapycnal upwelling along sloping topography has been shown to be an important component of the abyssal overturning circulation. Theoretical studies of mixing-driven upwelling have mostly relied on a time-averaged description of mixing acting on a mean stratification which ignores the intermittency of mixing. Typically, these studies prescribed a time-invariant turbulent diffusivity profile motivated by scenarios where tidal currents encounter gentle topography with small-scale corrugations, leading to subsequent propagation and breaking of internal waves. Here, a different scenario is considered where a tidal current interacts with smooth but steep topography, a case often encountered near continental margins and troughs. The performed nonhydrostatic simulations resolve both the strong oscillatory shear that develops along the steep critical topography and the associated mixing events. Strong diapycnal mixing is observed during the upslope phase of the tidal flow when both the near-boundary stratification and shear are enhanced. During the downslope phase, strong overturning events do develop, but they are associated with weak stratification and less efficient diapycnal mixing. These results highlight that the temporal evolution of both shear and stratification play a key role in setting when diapycnal mixing and water mass transformation occur along steep topography. In contrast, over gentle topography, tidal shears do not become sufficiently large to generate strong local mixing for typical oceanographic parameters. 

Abstract We discuss the emerging advances and opportunities at the intersection of machine learning (ML) and climate physics, highlighting the use of ML techniques, including supervised, unsupervised, and equation discovery, to accelerate climate knowledge discoveries and simulations. We delineate two distinct yet complementary aspects: (a) ML for climate physics and (b) ML for climate simulations. Although physics-free ML-based models, such as ML-based weather forecasting, have demonstrated success when data are abundant and stationary, the physics knowledge and interpretability of ML models become crucial in the small-data/nonstationary regime to ensure generalizability. Given the absence of observations, the long-term future climate falls into the small-data regime. Therefore, ML for climate physics holds a critical role in addressing the challenges of ML for climate simulations. We emphasize the need for collaboration among climate physics, ML theory, and numerical analysis to achieve reliable ML-based models for climate applications. 

Abstract Current eddy‐permitting and eddy‐resolving ocean models require dissipation to prevent a spurious accumulation of enstrophy at the grid scale. We introduce a new numerical scheme for momentum advection in large‐scale ocean models that involves upwinding through a weighted essentially non‐oscillatory (WENO) reconstruction. The new scheme provides implicit dissipation and thereby avoids the need for an additional explicit dissipation that may require calibration of unknown parameters. This approach uses the rotational, “vector invariant” formulation of the momentum advection operator that is widely employed by global general circulation models. A novel formulation of the WENO “smoothness indicators” is key for avoiding excessive numerical dissipation of kinetic energy and enstrophy at grid‐resolved scales. We test the new advection scheme against a standard approach that combines explicit dissipation with a dispersive discretization of the rotational advection operator in two scenarios: (a) two‐dimensional turbulence and (b) three‐dimensional baroclinic equilibration. In both cases, the solutions are stable, free from dispersive artifacts, and achieve increased “effective” resolution compared to other approaches commonly used in ocean models. 

Abstract Small-scale turbulent mixing drives the upwelling of deep water masses in the abyssal ocean as part of the global overturning circulation¹. However, the processes leading to mixing and the pathways through which this upwelling occurs remain insufficiently understood. Recent observational and theoretical work^2–5has suggested that deep-water upwelling may occur along the ocean’s sloping seafloor; however, evidence has, so far, been indirect. Here we show vigorous near-bottom upwelling across isopycnals at a rate of the order of 100 metres per day, coupled with adiabatic exchange of near-boundary and interior fluid. These observations were made using a dye released close to the seafloor within a sloping submarine canyon, and they provide direct evidence of strong, bottom-focused diapycnal upwelling in the deep ocean. This supports previous suggestions that mixing at topographic features, such as canyons, leads to globally significant upwelling^3,6–8. The upwelling rates observed were approximately 10,000 times higher than the global average value required for approximately 30 × 10⁶m³s⁻¹of net upwelling globally⁹. 

Abstract The Indo-Pacific Ocean appears exponentially stratified between 1- and 3-km depth with a decay scale on the order of 1 km. In his celebrated paper “Abyssal recipes,” W. Munk proposed a theoretical explanation of these observations by suggesting a pointwise buoyancy balance between the upwelling of cold water and the downward diffusion of heat. Assuming a constant upwelling velocity w and turbulent diffusivity κ , the model yields an exponential stratification whose decay scale is consistent with observations if κ ∼ 10 −4 m 2 s −1 . Over time, much effort has been made to reconcile Munk’s ideas with evidence of vertical variability in κ , but comparably little emphasis has been placed on the even stronger evidence that w decays toward the surface. In particular, the basin-averaged w nearly vanishes at 1-km depth in the Indo-Pacific. In light of this evidence, we consider a variable-coefficient, basin-averaged analog of Munk’s budget, which we verify against a hierarchy of numerical models ranging from an idealized basin-and-channel configuration to a coarse global ocean simulation. Study of the budget reveals that the decay of basin-averaged w requires a concurrent decay in basin-averaged κ to produce an exponential-like stratification. As such, the frequently cited value of 10 −4 m 2 s −1 is representative only of the bottom of the middepths, whereas κ must be much smaller above. The decay of mixing in the vertical is as important to the stratification as its magnitude . Significance Statement Using a combination of theory and numerical simulations, it is argued that the observed magnitude and shape of the global ocean stratification and overturning circulation appear to demand that turbulent mixing increases quasi-exponentially toward the ocean bottom. Climate models must therefore prescribe such a vertical profile of turbulent mixing in order to properly represent the heat and carbon uptake accomplished by the global overturning circulation on centennial and longer time scales. 

Abstract Small-scale mixing drives the diabatic upwelling that closes the abyssal ocean overturning circulation. Indirect microstructure measurements of in-situ turbulence suggest that mixing is bottom-enhanced over rough topography, implying downwelling in the interior and stronger upwelling in a sloping bottom boundary layer. Tracer Release Experiments (TREs), in which inert tracers are purposefully released and their dispersion is surveyed over time, have been used to independently infer turbulent diffusivities—but typically provide estimates in excess of microstructure ones. In an attempt to reconcile these differences, Ruan and Ferrari (2021) derived exact tracer-weighted buoyancy moment diagnostics, which we here apply to quasi-realistic simulations. A tracer’s diapycnal displacement rate is exactly twice the tracer-averaged buoyancy velocity, itself a convolution of an asymmetric upwelling/downwelling dipole. The tracer’s diapycnal spreading rate, however, involves both the expected positive contribution from the tracer-averaged in-situ diffusion as well as an additional non-linear diapycnal distortion term, which is caused by correlations between buoyancy and the buoyancy velocity, and can be of either sign. Distortion is generally positive (stretching) due to bottom-enhanced mixing in the stratified interior but negative (contraction) near the bottom. Our simulations suggest that these two effects coincidentally cancel for the Brazil Basin Tracer Release Experiment, resulting in negligible net distortion. By contrast, near-bottom tracers experience leading-order distortion that varies in time. Errors in tracer moments due to realistically sparse sampling are generally small (< 20%), especially compared to the O (1) structural errors due to the omission of distortion effects in inverse models. These results suggest that TREs, although indispensable, should not be treated as “unambiguous” constraints on diapycnal mixing. 

We numerically and theoretically investigate the Boussinesq Eady model, where a rapidly rotating density-stratified layer of fluid is subject to a meridional temperature gradient in thermal wind balance with a uniform vertically sheared zonal flow. Through a suite of numerical simulations, we show that the transport properties of the resulting turbulent flow are governed by quasigeostrophic (QG) dynamics in the rapidly rotating strongly stratified regime. The ‘vortex gas’ scaling predictions put forward in the context of the two-layer QG model carry over to this fully three-dimensional system: the functional dependence of the meridional flux on the control parameters is the same, the two adjustable parameters entering the theory taking slightly different values. In line with the QG prediction, the meridional heat flux is depth-independent. The vertical heat flux is such that turbulence transports buoyancy along isopycnals, except in narrow layers near the top and bottom boundaries, the thickness of which decreases as the diffusivities go to zero. The emergent (re)stratification is set by a simple balance between the vertical heat flux and diffusion along the vertical direction. Overall, this study demonstrates how the vortex-gas scaling theory can be adapted to quantitatively predict the magnitude and vertical structure of the meridional and vertical heat fluxes, and of the emergent stratification, without additional fitting parameters. 

Abstract ABSTRACT: The abyssal overturning circulation is thought to be primarily driven by small-scale turbulent mixing. Diagnosed watermass transformations are dominated by rough topography “hotspots”, where the bottom-enhancement of mixing causes the diffusive buoyancy flux to diverge, driving widespread downwelling in the interior—only to be overwhelmed by an even stronger up-welling in a thin Bottom Boundary Layer (BBL). These watermass transformations are significantly underestimated by one-dimensional (1D) sloping boundary layer solutions, suggesting the importance of three-dimensional physics. Here, we use a hierarchy of models to generalize this 1D boundary layer approach to three-dimensional eddying flows over realistically rough topography. When applied to the Mid-Atlantic Ridge in the Brazil Basin, the idealized simulation results are roughly consistent with available observations. Integral buoyancy budgets isolate the physical processes that contribute to realistically strong BBL upwelling. The downwards diffusion of buoyancy is primarily balanced by upwelling along the sloping canyon sidewalls and the surrounding abyssal hills. These flows are strengthened by the restratifying effects of submesoscale baroclinic eddies and by the blocking of along-ridge thermal wind within the canyon. Major topographic sills block along-thalweg flows from restratifying the canyon trough, resulting in the continual erosion of the trough’s stratification. We propose simple modifications to the 1D boundary layer model which approximate each of these three-dimensional effects. These results provide local dynamical insights into mixing-driven abyssal overturning, but a complete theory will also require the non-local coupling to the basin-scale circulation.