High Energy Astrophysics 4 JAN, 2015

Modeling the optical/UV polarization while flying around the tilted outflows of NGC 1068

Modeling the optical/UV polarization while flying around the tilted outflows of NGC 1068

Recent modeling of multi-waveband spectroscopic and maser observations suggests that the ionized outflows in the nuclear region of the archetypal Seyfert-2 galaxy NGC 1068 are inclined with respect to the vertical axis of the obscuring torus. Based on this suggestion, we build a complex reprocessing model of NGC 1068 for the optical/UV band. We apply the radiative transfer code STOKES to compute polarization spectra and images. The effects of electron and dust scattering and the radiative coupling occurring in the inner regions of the multi-component object are taken into account and evaluated at different polar and azimuthal viewing angles. The observed type-1/type-2 polarization dichotomy of active galactic nuclei is reproduced. At the assumed observer’s inclination toward NGC 1068, the polarization is dominated by scattering in the polar outflows and therefore it indicates their tilting angle with respect to the torus axis. While a detailed analysis of our model results is still in progress, we briefly discuss how they relate to existing polarization observations of NGC 1068.

💡 Research Summary

The paper addresses a long‑standing puzzle in active‑galaxy research: how the optical/UV polarization signatures of the archetypal Seyfert 2 galaxy NGC 1068 can be reconciled with recent multi‑waveband observations that suggest the ionized polar outflows are not aligned with the symmetry axis of the obscuring torus but are tilted by roughly 18–30°. To explore this, the authors construct a fully three‑dimensional radiative‑transfer model that incorporates four key components: (i) a central point source emitting a power‑law continuum (α = 1), (ii) a geometrically thick dusty torus (inner radius 0.1 pc, outer radius 0.5 pc, half‑opening angle 60°, optical depth τ≈750), (iii) an equatorial electron scattering region (a flared disk with τ≈1, extending from 0.01 to 0.1 pc), and (iv) bipolar polar outflows (inner radius 0.3 pc, outer radius 1.5 pc, half‑opening angle 40°, electron scattering optical depth τ≈0.03). Crucially, the polar outflows are rotated by 18° with respect to the torus axis, in line with maser and radio interferometric constraints.

The STOKES Monte‑Carlo code is employed to follow photon packets through this geometry, accounting for both Thomson (electron) scattering and Mie scattering by dust grains. The code records the Stokes parameters for each packet, enabling the construction of wavelength‑dependent polarization spectra (degree P(λ) and position angle ψ(λ)) as well as two‑dimensional polarization maps (Q and U images). Simulations are performed for a range of observer inclinations i (0°–90°) and azimuthal angles φ (0°, 90°, 180°, 270°). The inclination i ≈ 73° is identified as the most realistic line‑of‑sight for NGC 1068 based on previous interferometric studies.

Key results are as follows:

  1. Reproduction of the type‑1/type‑2 dichotomy – For face‑on views (i ≈ 0°), the equatorial electron scattering region dominates, yielding low polarization (P ≤ 1 %) with a position angle parallel to the torus plane (ψ ≈ 0°). For edge‑on views (i ≈ 73°), the polar outflows become the primary scattering sites; the resulting polarization is high (5–15 %) and the position angle is perpendicular to the torus axis (ψ ≈ 90°). This matches the classic observational distinction between Seyfert 1 (parallel) and Seyfert 2 (perpendicular) polarizations.

  2. Detection of the outflow tilt through azimuthal dependence – Because the outflows are tilted, the position angle ψ shows a modest φ‑dependence. At φ = 0° and 180° the angle stays close to 90°, whereas at φ = 90° and 270° it deviates by roughly ±10°. This subtle rotation reproduces the asymmetry seen in high‑resolution optical polarimetry of NGC 1068, where the northern cone appears slightly brighter and its polarization vector is not perfectly orthogonal to the radio jet.

  3. Wavelength‑dependent scattering contributions – In the far‑UV (λ < 300 nm) electron scattering dominates, giving the highest P and a stable ψ. In the near‑UV/optical (400–500 nm) dust (Mie) scattering becomes significant, causing a modest reduction in P and a rotation of ψ by about 5°. This reproduces the observed “polarization angle swing” across the Balmer jump reported in earlier spectropolarimetric studies.

  4. Polarization imaging – The simulated Q and U maps show strong, uniform polarization across the bipolar cones, while the torus shadow region exhibits negligible polarization. The tilt of the cones introduces a slight offset in the polarization vectors across the two sides, a feature that aligns with HST and VLT polarimetric imaging of NGC 1068.

The authors discuss model limitations: the torus and outflow density distributions are simplified, dust grain properties are fixed, and time‑dependent effects (e.g., variability of the central source) are not included. Nevertheless, the work demonstrates that a modest tilt of the polar outflows can explain several long‑standing observational anomalies without invoking exotic physics.

Future directions proposed include extending the model to X‑ray energies (where Compton scattering would dominate), incorporating clumpy torus structures, and performing a Bayesian inversion of high‑resolution polarimetric data from JWST and the VLTI to constrain the tilt angle more precisely. The authors conclude that optical/UV polarization, when interpreted with realistic three‑dimensional radiative‑transfer models, provides a powerful diagnostic of the inner geometry of AGN, and that NGC 1068 serves as a benchmark case for testing such models.