Towards realistic ocean flow decomposition using optimal balance with time-averaging
Silvano G. Rosenau 1, M. Chouksey 1,2, C. Eden 1 · 2025
Read Towards realistic ocean flow decomposition using optimal balance with time-averaging by Silvano G. Rosenau 1, M. Chouksey 1,2, C. Eden 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-11464.txt.
Introduction
Silvano Rosenau and co-authors present a 2025 paper on splitting ocean flow into slow and fast parts. Ocean currents mix a slow, balanced drift with fast waves riding on it. Separating them is essential for studying how waves are born from currents. Simple splits by size or speed are too crude. Optimal balance and normal mode methods handle nonlinear effects but worked only in idealized setups. Real coastlines, depth changes, and varying Coriolis force broke them.
The authors modify optimal balance with a time-averaging step that projects onto the slow part. The trick removes the need for the idealized assumptions. Ocean modelers get a decomposition that survives realistic geography. Spontaneous wave emission is a small energy leak with big climate model consequences. Cleaner splits sharpen estimates of that leak in real ocean simulations.
Catalog record
| CREATOR | Silvano G. Rosenau 1, M. Chouksey 1,2, C. Eden 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 74 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-11464
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Towards realistic ocean flow decomposition using optimal balance with time-averaging Silvano G. Rosenau 1, M. Chouksey 1,2, C. Eden 1 Abstract Decomposing oceanic and atmospheric flow fields into their slowly evolving balanced components and fast evolving wave components is essential for understanding processes like spontaneous wave emission. To study these processes, the decomposition into the linear geostrophic (slow) and non-geostrophic (fast) components is not precise enough. More precise methods, such as optimal balance and nonlinear normal mode decomposition, account for nonlinear effects, but their application has so far been limited to idealized model configurations that exclude irregular lateral boundaries, depth variations, a varying Coriolis parameter, or other environmental conditions. Here, we address these limitations by modifying the optimal balance method such that it is applicable to more realistic model setups. The modification employs a time-averaging procedure to project onto the linear geostrophic component, eliminating the need for a Fourier transformation as required in the original optimal balance method. We demonstrate analytically and experimentally that the new method converges to the original method when either the time-averaging period or the number of optimal balance iterations increases. We test the new method using a two-dimensional single-layer model and a three-dimensional non-hydrostatic model with varying initial conditions and Rossby numbers ranging from 0.03 to 0.5. In all tested configurations, the differences between balanced states obtained from the new method and those from the original method become exponentially small. Optimal balance with time-averaging thus comes forth as a promising new flow decomposition tool for complex flows. keywords Authors should not enter keywords on the manuscript, as these must be chosen by the author during the online submission process and will then be added during the typesetting process (see Keyword PDF for the full list). Other classifications will be added at the same time. † † affiliation: 1 Institut für Meereskunde, Universität Hamburg, Germany 2 Leibniz Institut für Ostseeforschung Warnemünde, Germany 1 Introduction Geophysical flows can be understood as a two timescale system consisting of a slow and a fast component. The need to decompose a flow into its slowly evolving balanced component and its fast wave components first emerged in geophysical research as a way to minimize artificial, strong, fast oscillations caused by gravity waves in weather forecast models. These fast oscillations occur when initializing numerical models with observations ( Temperton & Williamson, 1981 ) and degrade the quality of weather forecasts. Although that problem was resolved by other means than explicit flow decomposition ( Lynch & Huang, 1992 ) , there has recently been a resurgence of interest in flow decomposition methods, particularly concerning the determination of energy transfer between different scales and dynamical regimes like waves and geostrophically balanced flow. Energy is transferred from the atmosphere to the ocean predominantly by wind forcing. Once in the ocean, this energy tends to propagate towards larger scales within the inverse energy cascade ( Scott & Wang, 2005 ) . To maintain an equilibrium state, energy must be removed from the large scales, which are typically in geostrophic balance and evolve slowly. Several processes have been proposed for this, one of which is the spontaneous wave emission from the balanced flow (e.g. Vanneste, 2013 ; Chouksey et al., 2018 ). This loss of geostrophic balance on large scales would lead to an energy transfer to gravity waves, then within the wave field to a downscale energy transfer, and finally collapse to small-scale turbulence and molecular dissipation (e.g. Molemaker et al., 2005 ). In the atmosphere, the process of spontaneous wave emission is thought to be a significant source of gravity waves ( Hertzog et al., 2008 ; Plougonven & Zhang, 2014 ) . If dissipation by spontaneous wave emission plays a relevant role in the oceanic or atmospheric energy budgets, this process needs to be parameterized in general circulation models, as it might potentially be unresolved. However, before developing parameterizations, we first need to understand under which conditions waves are generated and how much energy is transferred. Answering these questions requires a precise flow decomposition method, which is the focus of this study. Several decomposition methods have been proposed, either by addressing nonlinearity directly, such as the nonlinear normal mode decomposition (NNMD) ( Machenhauer, 1977 ; Warn et al., 1995 ; Eden et al., 2019a ) or the optimal balance (OB) ( Masur & Oliver, 2020 ) modified from the optimal potential vorticity balance ( Viúdez & Dritschel, 2004 ) , or by simple temporal filtering, such as the digital filter initialization ( Lynch & Huang, 1992 ) or the particle-based Lagrangian filtering ( Shakespeare & Taylor, 2016 ) , which offer less accuracy. However, no decomposition method currently exists that simultaneously works in realistic model setups (with irregular boundaries, varying depth, and variable Coriolis parameter) and decomposes the flow with accuracy high enough to study spontaneous wave emission, like NNMD or OB. In this work, we present a modification of the OB method that can be used in models that include changes in depth and irregular lateral boundaries. With this modification, studying energy transfers between waves and balanced flow becomes possible in more realistic ocean models. In the following sections, we first introduce the system of equations and explain the concept of the slow (geostrophically balanced) manifold. Using that concept, in Section 3 , we introduce the OB method and explain how it works. We then present the novel OBTA acronym not introduced before method and examine its convergence behavior from an analytical perspective. Thereafter, we validate the convergence behavior in a numerical two-dimensional single-layer model and a three-dimensional non-hydrostatic model. Key conclusions are summarized in the final section. 2 Theory The concept of OB is based on the idea of a slow manifold. An intuitive understanding of the slow manifold helps in comprehending how and why these decomposition methods work. Therefore, we first introduce and discuss the slow manifold before detailing the decomposition methods in Section 3 . Although, for simplicity, we use a two-dimensional single-layer model (shallow water model) in this and the next section, the concepts are not limited to this model and can be readily extended to the three-dimensional primitive equations or the three-dimensional non-hydrostatic model used later in this work. 2.1 Shallow water model equations We consider the two-dimensional scaled equations for the flow of a single layer (shallow water) model in the double periodic domain \Omega=[0,2\upi]\times[0,2\upi] on the f -plane, given by: \partial_{t}\boldsymbol{z}=\mathcal{L}\bcdot\boldsymbol{z}+Ro\,\boldsymbol{N}(\boldsymbol{z}), (1) with: \boldsymbol{z}=\begin{pmatrix}u\\ v\\ h\end{pmatrix},\quad\mathcal{L}=\begin{pmatrix}0&f&-\partial_{x}\\ -f&0&-\partial_{y}\\ -c^{2}\partial_{x}&-c^{2}\partial_{y}&0\end{pmatrix},\quad\boldsymbol{N}(\boldsymbol{z})=\begin{pmatrix}-(u\partial_{x}+v\partial_{y})u\\ -(u\partial_{x}+v\partial_{y})v\\ -\partial_{x}(uh)-\partial_{y}(vh)\end{pmatrix} (2) The state vector \boldsymbol{z} contains the horizontal velocity components u and v and the layer thickness h . The linear operator \mathcal{L} contains the constant Coriolis parameter f and the Burger number c . The non-linear term \boldsymbol{N}(\boldsymbol{z}) contains the advection terms. The Rossby number Ro is a measure of the non-linearity of the system and is given by Ro=U/(fL) , where U is the typical velocity scale and L is the typical length scale of the system. We note that this system is identical to the single-layer model discussed by Chouksey et al. (2023, their Section 2.1, Eq. 2.3) The dynamics of a state vector \boldsymbol{z} for the linear system with Ro=0 can be described as a superposition of two inertia-gravity wave modes and