Statistics of an Unstable Barotropic Jet from a Cumulant Expansion

Low-order equal-time statistics of a barotropic flow on a rotating sphere are investigated. The flow is driven by linear relaxation toward an unstable zonal jet. For relatively short relaxation times, the flow is dominated by critical-layer waves. For sufficiently long relaxation times, the flow is turbulent. Statistics obtained from a second-order cumulant expansion are compared to those accumulated in direct numerical simulations, revealing the strengths and limitations of the expansion for different relaxation times.

💡 Research Summary

The paper investigates equal‑time low‑order statistics of a barotropic flow on a rotating sphere that is forced by a linear relaxation toward an unstable zonal jet. The relaxation time τ controls how strongly the jet is restored toward a prescribed profile; short τ corresponds to strong forcing and rapid relaxation, while long τ allows the flow to evolve more freely. In the short‑τ regime the dynamics are dominated by critical‑layer Rossby‑type waves that grow linearly on the jet and interact weakly with the mean flow. As τ increases, the linear restoring term weakens, non‑linear triadic interactions become dominant, the waves break, and the system transitions to a turbulent state with a broadband energy spectrum.

Two complementary methodologies are employed. First, direct numerical simulations (DNS) are performed using a high‑resolution spherical‑harmonic spectral solver (up to ℓ≈120) to integrate the full non‑linear barotropic vorticity equation with the relaxation term. Time‑averaged first‑order moments (the mean zonal jet) and second‑order cumulants (covariances of velocity fluctuations) are accumulated over thousands of planetary rotations to obtain statistically converged reference data. Second, a second‑order cumulant expansion (CE2) is derived by writing evolution equations for the mean flow ⟨u⟩ and the second cumulant C₂=⟨u′u′⟩ and closing the hierarchy by setting all third‑order and higher cumulants to zero. This closure yields a deterministic set of equations that can be integrated at a fraction of the computational cost of DNS.

The authors compare CE2 predictions with DNS results across a range of τ values. For τ≈0.1 (short relaxation), CE2 reproduces the mean jet profile, its width, and the phase structure of the critical‑layer waves with remarkable fidelity. The agreement is attributed to the fact that the flow is essentially linear; the second‑order statistics capture the dominant dynamics, and the neglected third‑order terms are indeed small. When τ is increased to order unity, the mean jet strength predicted by CE2 begins to be underestimated, and the asymmetry of the wave‑jet interaction is not fully captured. This regime marks the onset of appreciable non‑linear transfer, indicating that third‑order correlations start to play a role. For τ≥5 (long relaxation), the flow is fully turbulent. DNS shows a robust inverse cascade of kinetic energy, a pronounced high‑wavenumber tail in the spectrum, and a reduction of enstrophy consistent with strong non‑linear dissipation. CE2, however, fails to generate the high‑wavenumber energy, severely underestimates jet amplitude, and yields an overly smooth mean flow. The inability of CE2 to represent the turbulent cascade is a direct consequence of the truncation at second order, which eliminates the triadic interactions that drive the cascade.

The paper also examines the convergence properties of CE2. In the short‑τ limit the system rapidly reaches a statistically steady state independent of the initial condition. In the intermediate τ range, multiple quasi‑steady states can exist, leading to a dependence of the final CE2 solution on the initial mean flow; convergence can be slow and may require careful initialization. For long τ, CE2 either fails to converge or converges to an unphysical solution (e.g., an excessively flattened jet), underscoring the limitation of the second‑order closure in strongly non‑linear regimes.

In summary, the study demonstrates that a second‑order cumulant expansion provides an efficient and accurate statistical description of barotropic jet dynamics when the flow is dominated by linear wave mechanisms (short relaxation times). However, once non‑linear interactions become significant and turbulence develops, higher‑order cumulants (e.g., CE3 or CE3*) are required to capture the essential physics of energy transfer and jet maintenance. These findings have direct implications for the development of low‑cost statistical parameterizations in global climate and atmospheric models: CE2 can be safely employed in regimes where linear wave–mean flow coupling prevails, but modelers must incorporate higher‑order closures or hybrid approaches to handle fully turbulent regimes. The paper thus clarifies the domain of applicability of low‑order statistical closures and points toward future work on extending cumulant expansions to include essential non‑linear dynamics.