Polarised Radiative Transfer Calculations

• Polarised radiative transfer (PRT) is a framework that uses the full Stokes vector to model electromagnetic radiation, incorporating scattering, absorption, and magneto-optical effects.
• The methodology involves solving coupled first-order ODEs with advanced numerical techniques like DELO-Hermitian and Fourier decomposition, ensuring high accuracy in complex media.
• PRT calculations enable precise modeling of spectral line formation, magnetic field diagnostics, and radiative processes in both astrophysical and laboratory contexts.

Polarised radiative transfer (PRT) calculations describe the transport, scattering, emission, and absorption of electromagnetic radiation in astrophysical and laboratory environments, accounting rigorously for the vector (polarisation) nature of the radiation field. Unlike scalar radiative transfer, PRT is governed by a system of coupled first-order ODEs for the Stokes parameters $(I,Q,U,V)$, subject to absorption, dichroism, Faraday rotation, and magneto-optical effects, as well as nontrivial emission and scattering source terms. These equations are central to modeling spectral line formation, synchrotron emission, the Hanle and Zeeman effects, depolarisation phenomena, and the inference of magnetic-field geometries in a broad range of astrophysical and atmospheric contexts.

1. Fundamental Equations and Formalism

The vector radiative transfer equation for the Stokes vector $\mathbf{I}(s)\equiv (I,Q,U,V)^{\mathrm{T}}$ is

$$\frac{d\mathbf{I}}{ds} = -\mathbf{K}(s)\,\mathbf{I}(s) + \boldsymbol{\epsilon}(s)$$

where $\mathbf{K}(s)$ is the $4\times4$ propagation matrix and $\boldsymbol{\epsilon}(s)$ is the emission term. The most general form of $\mathbf{K}$ includes absorption ($\eta_I$), dichroism ($\eta_Q,\eta_U,\eta_V$), and dispersion (magneto-optical $\,\rho_Q,\rho_U,\rho_V$) coefficients (Janett et al., 2018): $$\mathbf{K} = \begin{pmatrix} \eta_I & \eta_Q & \eta_U & \eta_V \ \eta_Q & \eta_I & \rho_V & -\rho_U \ \eta_U & -\rho_V & \eta_I & \rho_Q \ \eta_V & \rho_U & -\rho_Q & \eta_I \end{pmatrix}$$ These coefficients are physically derived from the microphysics of the plasma (e.g., densities, magnetic fields, line profile functions) and require evaluation of the local conditions at each spatial point, frequency, and direction.

Scattering source terms are often written as integrals over frequencies and directions, subdivided into partial-frequency-redistribution (PRD) and complete redistribution (CRD) terms for spectral lines, or as convolution with phase matrices for continuous media and dust (Ballester et al., 2022, Anusha et al., 2013).

2. Key Physical Effects: Scattering, Magnetic Fields, and Redistribution

a) Scattering and Emission

The source term for line scattering in an NLTE, polarised, turbulent or magnetised medium is typically expressed as: $$\epsilon_i(\nu, \Omega) = k \int d\nu' \oint \frac{d\Omega'}{4\pi} \mathcal{R}_{ij}(\nu', \Omega'; \nu, \Omega) I_j(\nu', \Omega') + \epsilon^{\text{th}}_i$$ where $\mathcal{R}$ is the PRD redistribution matrix, encoding angular and frequency correlations, quantum interference, Hanle/Zeeman, and often hyperfine structure terms (Riva et al., 27 May 2025, Ballester et al., 2022).

b) Polarisation Mechanisms

Line polarisation arises predominantly from anisotropic scattering and is further modulated via the Hanle (magnetic-induced depolarisation/rotation in weak fields), Zeeman (splitting and polarisation signatures in strong fields), and incomplete Paschen-Back effects. Magneto-optical effects such as Faraday rotation (linear $\leftrightarrow$ linear, $Q \leftrightarrow U$) and Faraday conversion (linear/circular interconversion, $Q/U \leftrightarrow V$) are encoded in the non-diagonal elements of $\mathbf{K}$ (Janett et al., 2018, Ballester et al., 2022). Proper treatment of PRD and J-state interference is essential in strong resonance lines (Riva et al., 27 May 2025).

c) Partial Frequency Redistribution and J-state Interference

PRD describes the correlation between the frequencies of incoming and outgoing photons in scattering events, particularly critical in strongly scattering lines with subordinate levels (e.g., Mg II h&k, H I Ly$\alpha$). The data-rich approach couples the angular redistribution with quantum-number interference (J-state or hyperfine splitting) in the presence of magnetic fields (Ballester et al., 2022, Riva et al., 27 May 2025).

3. Numerical Methods and Quadrature Strategies

a) Formal Solvers

Formal integration of the coupled ODE system employs a variety of schemes. High-fidelity solvers include:

• Cubic Hermitian (DELO-Hermitian): Unconditionally A-stable, fourth-order, robust for stiff regions (Janett et al., 2017).
• Quadratic/Cubic DELO-Bézier: Affine in the unknown, second/third-order, monotonic, free from artefactual oscillations in strong gradients (Rodríguez et al., 2012).
• Explicit/Implicit Runge–Kutta and Adams–Moulton integrators: Suitable for mild stiffness; bounded absolute stability in optically thick layers necessitates hybrid switching to A-stable methods (Janett et al., 2018, Janett et al., 2017).
• Adaptive/Pragmatic Solvers: Automatic step-size or method switching depending on local stiffness (as measured by spectrum of $-\mathbf{K}$), L-stability for extreme damping (Janett et al., 2018).

b) Angular and Spectral Integration

Efficient, highly accurate quadrature for integrating the radiation field tensors and source integrals is crucial in 3D or NLTE problems:

• Near-optimal quadratures provide up to $30\%$ reduction in ray directions for a given multipole accuracy, matching all moments up to arbitrary tensor rank $L$ (Stepan et al., 2020). Construction follows a moment-matching procedure supplemented by symmetry constraints and least-squares minimisation to machine precision.
• Fourier and Spherical Harmonic Decomposition: Decouple azimuthal and polar dependencies, allowing for block-diagonalization and reduction in computational dimensionality (Corbett, 2019, Anusha et al., 2013, Tapimo et al., 2018).
• Azimuthal/Fourier Expansion: For angle-dependent PRD, decomposes scattering integrals into modes labeled by $k$, solving smaller coupled systems (Anusha et al., 2013, Riva et al., 27 May 2025).

c) Preconditioning and Iterative Solution

Matrix-free, physics-based preconditioners using angle-averaged kernels and block Jacobi/diagonal approximations dramatically accelerate Krylov iterations in high-dimensional (e.g., $N \sim 10^7$) RT problems with PRD and magnetic fields (Riva et al., 27 May 2025).

4. Applications in Astrophysical and Laboratory Contexts

a) Stellar and Planetary Atmospheres

• Modeling NLTE line polarisation in solar and stellar atmospheres, with detailed treatments of PRD, Hanle/Zeeman effects, and 3D radiative transfer (Riva et al., 27 May 2025, Ballester et al., 2022, Anusha et al., 2013).
• Disc-resolved and phase-resolved polarisation of planetary atmospheres and exoplanets, including multiple scattering from clouds/hazes and full-disk integration with DOM schemes (Bailey et al., 2018, Wu et al., 2024).

b) Cosmic Magnetism and Galactic Synchrotron

• Full-Stokes modeling of Galactic synchrotron emission, Faraday rotation, and depolarisation within sophisticated MHD-based galaxy simulations (Reissl et al., 2019).
• Fluctuation analysis of Faraday rotation for diagnosing interstellar and intracluster magnetic fields, requiring proper PRT modeling to avoid biases from density and turbulence statistics (On et al., 2019).

c) General Relativistic Regimes

• Covariant PRT along null geodesics in arbitrary metrics for black hole accretion flows, jets, and strongly curved spacetimes, with explicit parallel transport of the polarisation basis, and rigorous handling of spacetime and plasma-induced effects (Shcherbakov et al., 2010, Pihajoki et al., 2016, Moscibrodzka, 2019, Dexter, 2016).
• Monte Carlo and ray-tracing schemes with invariant Stokes weighting, extending to full multifrequency Compton-polarisation transport (Moscibrodzka, 2019).

d) Laboratory and Oceanic Applications

• Analytic or numerically-exact solutions in homogeneous and inhomogeneous water bodies (e.g., ocean optics, remote sensing), exploiting azimuthal decoupling and Fourier spectral reduction for efficient computation (Corbett, 2019).
• Benchmarks and high-precision solutions in 1D slab geometry for Rayleigh and combined isotropic–polarising media via doubling–adding with convergence acceleration (Ganapol, 2017).

5. Advanced Theoretical Extensions and Structure-Preserving Methods

• Graded-index media: Generalisations of the VRTE to inhomogeneous refractive index environments introduce nontrivial ray curvature and geometric rotations (torsion/Rytov effect), requiring extra precession terms in the evolution of $(Q,U)$ (Zhao et al., 2011).
• Metriplectic formulations: A Hamiltonian–dissipative (metriplectic) structure for the PRT equations guarantees strict thermodynamic consistency (energy conservation, entropy monotonicity), via antisymmetric Poisson and symmetric metric brackets, and admits structure-preserving integrators for long-term stable computation (Bosboom et al., 2023).
• Integration in multidomain geometries: Discrete spherical harmonics (DSHM) and Chebyshev spectral representations enable global, high-accuracy solutions in structured multidimensional media (Tapimo et al., 2018).

6. Numerical Stability, Benchmarks, and Practical Recommendations

• High-order A-stable methods, particularly cubic Hermitian and DELO–Bézier schemes, are preferred for problems exhibiting a mixture of optically thin and thick regions, strong velocity or magnetic gradients, or when positivity and monotonicity of interpolants are critical (Janett et al., 2017, Rodríguez et al., 2012).
• Use of optical depth as an integration variable reduces numerical stiffness. Step sizes should adapt to local eigenvalues of $-\mathbf{K}$, with method switching per cell as necessary (Janett et al., 2018).
• Near-optimal angular quadratures and Fourier-mode decompositions yield direct computational savings in multidimensional, non-LTE problems (Stepan et al., 2020, Anusha et al., 2013).
• Extensive benchmarks against analytic and semi-analytic solutions in 1D, water bodies, and GRMHD disks validate all leading numerical frameworks to high (order $10^{-7}$) accuracy (Ganapol, 2017, Corbett, 2019, Moscibrodzka, 2019).

7. Outlook: Future Directions and Ongoing Challenges

• Extension of efficient, scalable polarised RT solvers—including full PRD, J-state interference, and 3D MHD—remains an active field, with emphasis on parallelized, matrix-free methods and adaptive gridding (Riva et al., 27 May 2025).
• Incorporation of more complex microphysics, e.g. partial coherence in continua, higher-order quantum interference, and non-LTE population coupling to polarisation, is required for precision modeling of increasingly detailed spectropolarimetric observations (Ballester et al., 2022).
• Structure-preserving numerical schemes (metriplectic/GREENIC) offer a path toward rigorously stable, long-time integration in both classical and general-relativistic PRT (Bosboom et al., 2023).
• Rigorous frameworks for linking observed polarised signatures to magnetic-field statistics—including proper treatment of statistical anisotropy and density PDFs—are critical to unambiguous inference from polarimetric data (On et al., 2019).

Polarised radiative transfer calculations are thus foundational to precision astrophysical inference, laboratory plasma diagnostics, and planetary or atmospheric characterization, demanding ongoing advances in mathematical formulation, numerical algorithms, and integration with high-resolution MHD and kinetic models.