Fast-Mode Compressible Turbulence

• Fast-mode compressible turbulence consists of nearly isotropic, compressive magnetosonic fluctuations that differ from Alfvén and slow modes by their polarization and geometric properties.
• Mode decomposition through eigenanalysis and spatiotemporal diagnostics reveals regime-dependent energy fractions and spectral slopes, highlighting distinct behaviors in MHD and kinetic plasma simulations.
• Observational signatures, such as isotropy in synchrotron emission and efficient particle acceleration, emphasize the role of fast modes in damping, transport, and cosmic-ray energization.

Fast-mode compressible turbulence is the part of magnetized turbulence carried by fast magnetosonic fluctuations, conventionally distinguished from Alfvén and slow modes by their compressive polarization, broader propagation-angle support, and much weaker scale-dependent anisotropy. In current literature it is not treated as a universal cascade with fixed spectral and geometric properties, but as a regime-dependent component whose measured energy fraction, wave character, and dissipation depend strongly on forcing, plasma beta, magnetization, partial ionization, and on whether diagnostics are based on instantaneous spatial projections or on full spatiotemporal spectra (Sebastian et al., 22 Apr 2026, Gan et al., 2022, Zhao et al., 9 Mar 2026).

1. Modal definition and decomposition framework

The standard theoretical framework begins with the decomposition of compressible MHD fluctuations into Alfvén, slow, and fast eigenmodes. In Fourier-space formulations used for compressible MHD, the fast displacement direction is written as

$$\hat{\zeta}_{\rm f} \propto (1+\alpha + \sqrt{D})(k_{\perp}\hat{\bm k}_{\perp}) +(-1 + \alpha+ \sqrt{D})(k_{\parallel}\hat{\bm k}_{\parallel}),$$

with

$D=(1+\alpha)^2-4 \alpha \cos^2\theta,$

while the Alfvén mode is transverse and the slow mode occupies the complementary compressive branch. This eigenvector-based projection underlies both synthetic-observation studies and transport calculations in compressible MHD (Wang et al., 2022, Zhang et al., 2021).

A major recent extension is the application of this mode picture to fully kinetic, relativistic, collisionless plasma turbulence. In a 3D PIC simulation of a magnetically dominated pair plasma, the decomposition is performed in the local mean-field frame, with

$$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$

and the second-order velocity structure function

$$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$

This is presented as the first application of the Cho–Lazarian mode decomposition to fully kinetic collisionless relativistic turbulence (Sebastian et al., 22 Apr 2026).

In relativistic MHD, the fast-mode polarization itself is modified by magnetization. In the high-$\sigma$, finite-temperature limit, the fast displacement acquires a finite component parallel to the background magnetic field rather than remaining purely perpendicular, which the relativistic literature identifies as a central reason for stronger Alfvén–fast coupling than in non-relativistic low-$\beta$ turbulence (Takamoto et al., 2017, Takamoto et al., 2016).

2. Geometry, anisotropy, and spectral behavior

Across a wide range of studies, the most stable qualitative result is the geometric distinction between anisotropic Alfvén/slow turbulence and nearly isotropic fast turbulence. In the fully kinetic relativistic PIC study, Alfvén and slow modes follow Goldreich–Sridhar-like anisotropy,

$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$

whereas fast modes remain nearly isotropic at all scales (Sebastian et al., 22 Apr 2026). Closely related structure-function and transport analyses in compressible MHD summarize the same contrast as

$l_{\parallel} \propto l_{\perp}^{2/3}$

for Alfvén and slow modes, but

$l_{\parallel} \propto l_{\perp}$

for fast modes (Gao et al., 2023, Zhang et al., 2021). In relativistic MHD simulations, fast-mode eddies are likewise reported as nearly isotropic, with

$r_{||} \propto r_{\perp},$

independent of scale (Takamoto et al., 2016).

Synthetic synchrotron diagnostics recover the same distinction observationally. The quadrupole-to-monopole ratio $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$0 increases with viewing angle for Alfvén and slow modes, while for fast mode $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$1 and varies only weakly with angle; contour maps remain close to circular over the explored line-of-sight geometries (Wang et al., 2022). In magnetosheath spatiotemporal spectra, the fast-mode power spectrum is also close to isotropic, with fitted slopes

$D=(1+\alpha)^2-4 \alpha \cos^2\theta,$2

supporting an approximately isotropic fast-mode cascade over the observed inertial range (Zhao et al., 9 Mar 2026).

Quantitatively, however, the spectral slope of fast-mode turbulence is not unique. Reported inertial-range behaviors include $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$3 for fast modes in an MHD simulation and $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$4 in a kinetic PIC simulation, the latter described as reminiscent of acoustic turbulence (Gan et al., 2022, Sebastian et al., 22 Apr 2026). Other MHD and RMHD studies report $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$5, $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$6, or spectra tending toward $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$7 in supersonic or damping-modified regimes (Gao et al., 2023, Hu, 14 Dec 2025, Kowal et al., 2010). In high-$D=(1+\alpha)^2-4 \alpha \cos^2\theta,$8 RMHD, the fast-mode spectrum is $D=(1+\alpha)^2-4 \alpha \cos^2\theta,$9 for $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$0 and steepens to $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$1 for $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$2 (Takamoto et al., 2016). A deliberately isolated 2D relativistic fast cascade gives $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$3 in the weakly driven regime and $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$4 in the strong shock-driven regime (Ugarov et al., 5 Apr 2026). The literature therefore supports isotropy more robustly than any single universal spectral index.

3. Energy content and intermode coupling

The energetic importance of fast modes is strongly regime dependent and is one of the main points of divergence between classical compressible MHD and relativistic collisionless turbulence. In 3D compressible MHD with nearly incompressible or magnetically driven forcing, a stringent spatiotemporal classification finds that the fast-wave fraction is essentially negligible: $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$5 in Run A and $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$6 in Run B. Even under highly compressible driving it reaches only $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$7, while about $$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$8–$$\mathbf{B}_{\ell} = \tfrac{1}{2}\big[\mathbf{B}(\mathbf{r}_1)+\mathbf{B}(\mathbf{r}_2)\big], \qquad \hat{\mathbf{z}} = \mathbf{B}_{\ell}/|\mathbf{B}_{\ell}|,$$9 of the total fluctuation power is classified as non-wave rather than Alfvén, slow, or fast (Gan et al., 2022). The same work shows that snapshot-only mode decomposition can overestimate fast content, returning $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$0, $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$1, and $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$2 for Runs A, B, and C, respectively, after excluding injection-scale power and degenerate regions (Gan et al., 2022).

By contrast, in fully kinetic relativistic turbulence the fast-mode share is substantially larger. The PIC simulation reports

$$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$3

compared with an MHD run giving

$$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$4

The fast fraction is therefore more than twice as large in the collisionless relativistic case, which the authors interpret as evidence for stronger coupling between Alfvén and fast modes in relativistic magnetized turbulence (Sebastian et al., 22 Apr 2026).

Relativistic MHD simulations make this coupling explicit in terms of magnetization. For $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$5,

$$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$6

with a value around $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$7, whereas for $$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$8,

$$SF_v(r_\|,r_\perp) = \big\langle \,(\mathbf{v}(\mathbf{r}_2)-\mathbf{v}(\mathbf{r}_1))^2 \,\big\rangle.$$9

A fitted form in isothermal RMHD is

$\sigma$0

with $\sigma$1 (Takamoto et al., 2016, Takamoto et al., 2017). Dedicated mode-conversion runs at $\sigma$2 further give steady partitions $\sigma$3 for initially Alfvénic turbulence and $\sigma$4 for initially fast-mode turbulence, demonstrating strong two-way conversion (Takamoto et al., 2017).

Not all compressible environments amplify fast modes in this manner. In partially ionized two-fluid turbulence, the fast-mode energy fraction remains near $\sigma$5 across strong and weak neutral–ion coupling, even though its spectrum steepens from approximately $\sigma$6 toward $\sigma$7 in the damping regime (Hu, 14 Dec 2025). In compressible MHD simulations used for cosmic-ray transport, the fast-mode fraction increases with $\sigma$8 but becomes insensitive to $\sigma$9, saturating at about $\beta$0 at fixed $\beta$1 (Hu et al., 2021). The cumulative picture is therefore not that fast modes are generically dominant, but that their energetic share is highly sensitive to magnetization, kinetics, and diagnostic method.

4. Wave character, spatiotemporal diagnostics, and the non-wave controversy

A central methodological issue is whether a fluctuation that projects onto the fast eigenvector is actually a propagating fast wave. The strongest argument against equating the two comes from spatiotemporal 4D-FFT analysis, in which spectral power is counted as fast only if it lies within a tolerance band around the theoretical fast-mode dispersion surface. In compressible MHD this procedure shows that most spatially “fast-like” power does not satisfy the fast-wave dispersion relation; the dominant component is non-wave, low-frequency structure (Gan et al., 2022). This directly undercuts the common simplification that compressible fluctuation power can be identified with fast-mode power.

Mode-resolved observations in Earth’s magnetosheath reinforce that fast modes are dynamically distinct from slow compressive turbulence. A multi-spacecraft polarization-based decomposition recovers full $\beta$2 spectra and finds that fast modes retain narrow peaks near $\beta$3 with only modest nonlinear broadening, whereas slow modes evolve from wave-like peaks to broad low-frequency continua as nonlinearity increases (Zhao et al., 9 Mar 2026). In this operational sense, fast modes remain weakly turbulent over the observed inertial range, while slow modes undergo a weak-to-strong transition.

Fully kinetic relativistic simulations in 2D show that a genuine fast-magnetosonic cascade can also exist as a weak wave-turbulence regime. With compressive in-plane forcing and out-of-plane $\beta$4, spatiotemporal power concentrates along the fast-mode dispersion relation for weak driving, but crosses into irregular shock-like dynamics as the driving amplitude increases (Ugarov et al., 5 Apr 2026). A related controversy concerns whether wave steepening destroys the fast cascade. Hybrid and PIC simulations of fast-mode turbulence argue that it does not: raw Fourier spectra are steepened by phase steepening, but structure-function analysis recovers an underlying cascade consistent with $\beta$5, while damping at MHD scales remains well described by linear transit-time damping theory (Hou et al., 5 Aug 2025). This suggests that strong nonlinearity and wave-based dissipation remain compatible within fast-mode turbulence rather than being mutually exclusive.

5. Collisionless damping, kinetic-scale modification, and relativistic effects

Kinetic physics modifies fast-mode turbulence in ways not captured by ideal MHD. In magnetosheath observations, collisionless damping introduces a truncation scale $\beta$6 defined by

$\beta$7

Above this scale, fast modes show weak, scale-independent anisotropy; below it, anisotropy strengthens with increasing wavenumber and the fast-mode fraction decreases with increasing $\beta$8 and propagation angle $\beta$9 (Zhao et al., 2023). The same study reports strong correlations between anisotropy and damping strength, with the anisotropy ratio $$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$0 correlating with $$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$1 at approximately $$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$2 (Zhao et al., 2023).

Kinetic simulations of decaying fast-mode turbulence identify transit-time damping as a quantitatively predictive dissipation mechanism even in a strongly nonlinear state. In the PIC case, the 2D power spectral density shows a pronounced angular dip near

$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$3

where the TTD rate peaks for the chosen parameters. At this strongest-damping angle the inertial range breaks near

$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$4

whereas in weakly damped directions the cascade persists to

$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$5

The same work interprets the truncation as damping-limited rather than as failure of cross-scale energy transfer (Hou et al., 5 Aug 2025).

In relativistic collisionless pair-plasma turbulence, an additional kinetic complication is the growth of thermal fluctuations near the electron skin depth $$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$6. These thermal fluctuations flatten the total velocity structure function, weaken apparent anisotropy, and reduce dynamic alignment; the alignment angle in the kinetic range approaches

$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$7

corresponding to effectively random orientation (Sebastian et al., 22 Apr 2026). In the dedicated 2D relativistic fast-cascade study, the fast branch itself evolves from a low-$$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$8 relativistic fast magnetosonic mode to an electromagnetic-wave limit at $$SF_v(r_\perp)\propto r_\perp^{2/3}, \qquad SF_v(r_\|)\propto r_\|^{1},$$9, again showing that the kinetic continuation of fast-mode turbulence is not a simple extrapolation of fluid MHD (Ugarov et al., 5 Apr 2026).

6. Observational signatures and astrophysical implications

Fast modes have a distinctive observational signature in synchrotron fluctuation statistics. The normalized correlation function and quadrupole-to-monopole ratio recover the expectation that Alfvén and slow modes become more anisotropic as the viewing angle increases, while the fast-mode contribution remains close to isotropic and nearly unchanged with angle (Wang et al., 2022). This makes weak or angle-insensitive anisotropy in synchrotron total or polarization intensity a practical indicator of a significant fast-mode contribution.

Solar-wind observations present a more ambiguous picture. Statistical analyses using Parker Solar Probe, Solar Orbiter, and Wind find that anti-correlated density and magnetic-pressure fluctuations, consistent with slow modes, dominate the compressible budget, exceeding $l_{\parallel} \propto l_{\perp}^{2/3}$0 in both Alfvénic and non-Alfvénic wind and exceeding $l_{\parallel} \propto l_{\perp}^{2/3}$1 in slow/non-Alfvénic wind. A correlated fast-mode-like component is present, but it is a minority contribution and is not reproduced by either linear MHD fast-mode theory or the tested nonlinear forced-compressible model (Gonzalez et al., 19 Feb 2026). Other fast-solar-wind work shows that compressibility can amplify the cascade rate by a factor of $l_{\parallel} \propto l_{\perp}^{2/3}$2 to $l_{\parallel} \propto l_{\perp}^{2/3}$3 in about $l_{\parallel} \propto l_{\perp}^{2/3}$4 of samples, but that result is framed as a property of the exact compressible energy flux rather than as evidence for fast-mode dominance (Banerjee et al., 2016). Likewise, 3D fast-solar-wind simulations driven by outward Alfvén waves conclude that compressibility is crucial through parametric decay instability and reflection, yet the developed turbulence remains imbalanced, anisotropic, and primarily Alfvénic rather than fast-mode dominated (Shoda et al., 2019).

The astrophysical importance of fast-mode turbulence is most pronounced in particle transport and acceleration. Test-particle simulations in compressible MHD repeatedly identify fast modes as efficient scatterers and accelerators because of their isotropy. In one study, particle acceleration is dominated by the fast mode in high-$l_{\parallel} \propto l_{\perp}^{2/3}$5 turbulence and by fast plus slow modes in low-$l_{\parallel} \propto l_{\perp}^{2/3}$6 turbulence, while magnetosonic modes remain central to diffusion and scattering even during acceleration (Gao et al., 2023). Another finds that fast mode dominates particle acceleration especially in super-Alfvénic and supersonic turbulence, with maximum acceleration-rate spectra following $l_{\parallel} \propto l_{\perp}^{2/3}$7 for fast modes versus $l_{\parallel} \propto l_{\perp}^{2/3}$8 for Alfvén and slow modes (Zhang et al., 2021). A complementary transport study identifies fast modes as the main agent for pitch-angle scattering, while Alfvénic turbulence controls perpendicular superdiffusion; in that framework, the suppression of diffusion in supersonic molecular clouds is attributed primarily to changes in $l_{\parallel} \propto l_{\perp}^{2/3}$9 rather than $l_{\parallel} \propto l_{\perp}$0 because the fast-mode fraction increases with $l_{\parallel} \propto l_{\perp}$1 but becomes insensitive to $l_{\parallel} \propto l_{\perp}$2 (Hu et al., 2021).

Taken together, these results establish fast-mode compressible turbulence as a distinct but non-uniform component of plasma turbulence: nearly isotropic in geometry, often minor in true wave power under ordinary MHD driving, but capable of becoming energetically enhanced and strongly coupled to Alfvénic turbulence in relativistic magnetically dominated plasmas, and disproportionately important for damping, transport, and stochastic particle energization (Gan et al., 2022, Sebastian et al., 22 Apr 2026, Takamoto et al., 2017).