Authors: E. Moy and B. Rocca-Volmerange
First Author’s Institution: Institut d’Astrophysique de Paris, 98 bis boulevard Arago, 75014 Paris, France; Max-Planck-Institut für extraterrestrische Physik, Postfach 1312, 85741 Garching, Germany
Status: Published in Astronomy & Astrophysics. Open access on arXiv.
A Tale of Two Jets
At the core of each Active Galactic Nucleus (AGN), matter funnels onto a dense supermassive black hole (SMBH). This cosmic spiral (accretion) can then interact with the magnetic field of the SMBH and produce sprawling bi-polar radio jets (i.e. synchrotron radiation), which emanate perpendicular to the galactic center. This powerful emission is often used to categorize their hosts as “radio galaxies”.
One of the most direct consequences of AGN activity is the ionization of neighboring gas. AGN photoionization, the process by which the AGN’s ionizing photons strike nearby atoms or molecules and strip them of their electron(s), has traditionally been suspected of producing ionized ultraviolet and optical emission lines near AGNs. On the other hand, shock waves have also been shown to play a critical role in this process. In particular, electron temperatures in the spectra of some AGN have been found to be ~ 20,000 Kelvin, a value higher than expected from pure AGN photoionization and more in line with shock waves propagating in the galaxy – which arise from the strong interaction between AGN winds and jets with the interstellar medium. To measure this, astronomers typically used the emission line ratio [OIII] λ4363 / [OIII] λ5007, which serves as a proxy for electron temperature! Is it possible, then, that photoionization and shockwaves can co-ionize near an AGN? If so, when does one dominate over the other?
A Closer Look
To tackle these inquiries, the authors of today’s paper systematically compare nuclear (only the central regions), extended (only off-nuclear regions), and spatially integrated (both nuclear and off-nuclear regions) emission line spectra for 369 radio galaxies with CLOUDY photoionization models and MAPPINGIII shock models. In doing so, they constrain the impact of both ionization sources on the radio sources they inspect.
For the photoionization models, they primarily explore the ionization parameter, U. This parameter measures the ratio of ionizing photon flux to the gas density and broadly varies between -4 ≥log(U) ≥ -1 in their models. For the shock models, they consider a wide range of velocities between 100 and 1000 km s-1. Figure 1 displays one of the results from their analysis using two key emission-line ratios, [OIII] λ5007 / Hβ λ4861 and [NII] λ6584 / Hα λ6563, which are highly ionization-sensitive.
Figure 1: Comparisons between the predictions of two-component photoionization + shocks models (thick solid lines) and observational data. The axes represent the ionization-sensitive ratios [OIII] λ5007 / Hβ λ4861 and [NII] λ6584 / Hα λ6563. The values of log U and v are indicated on the plot. Two pure photoionization model sequences (-4 ≤ log(U) ≤ -1), with spectral index (a measure of the amount of flux at a given frequency), α = -1 (thin solid line) and α = -1.5 (dotted line), are shown for comparison. Filled squares indicate emission-line ratios for spatially integrated (nuclear+extended) emission-line regions, filled triangles for nuclear regions only, and open circles for extended regions only. (Figure 1 in the paper.)
From Figure 1, they show that the AGN photoionization and shock sequences (thick solid lines) are very similar to the pure AGN photoionization sequence (thin line). This exemplifies that a variation of photoionization and shocks can mimic changes to the ionization parameter U. In fact, the pure photoionization lines (α = -1 and α = -1.5) actually underpredict the observational data (likely due to a lack of sufficient high-energy photons that correspond to these spectral indices). Comparatively, the photoionization and shock sequence with log (U) = -1 and v = 300 km s-1 provides a better fit for the data.
Moreover, most data points from the radio galaxies in Figure 1 are located above log([OIII λ5007 / Hꞵ λ4861) = 0.5, and only five show low ionization (log([OIII] λ5007 / Hꞵ λ4861) ∼ 0). For these low ionization sources, there are no nuclear regions (filled triangles) present. The authors reason that this low ionization region corresponds to zones of pure photoionization and interpret this as compelling evidence that shocks mainly occur far from the nucleus and work in tandem with photoionization!
The authors then analyze another ionization-sensitive ratio, [OII] λ3727/ [OIII] λ5007, as a function of the size of radio emission in their sample (i.e. radio size). From this investigation, they determine that compact sources (≤ 2 kpc) are, on average, dominated by photoionization. Shock contributions then intensify with radio size before becoming dominated by photoionization again at large radii. They posit that photoionization is the primary ionization mechanism within close proximity to the AGN, until the cocoon-like surface area of the shock wave expands and shock ionization takes over. Then, at extended distances (D ≥ 200 kpc), the shock front passes beyond the material surrounding the AGN and photoionization reigns supreme again. A beautiful dance, indeed.
Today’s authors reveal the exciting dynamics of AGN emission-line regions. While previous studies report photoionization to be the dominant AGN ionization mechanism, we now see new evidence that shocks may also play a vital role in the AGN ionization process – likely due to interactions between radio jets and the surrounding interstellar medium. The authors used a large sample of radio galaxies, with the help of models and prominent emission-line ratios, to show that photoionization and shock models can provide better fits to their data when compared to pure photoionization. This suggests that the two ionization sources coexist, with photoionization dominating closest to the AGN and shocks at intermediate distances (before photoionization takes over again further out). Looking forward, it will be intriguing to see how similar models with differing inputs (e.g., density, metallicity, magnetic fields) compare. In addition, using integral field spectroscopy (e.g., MaNGA), which can provide enhanced spatial resolution of AGN emission-line regions and their host galaxies, will also be beneficial in future studies.
Today’s paper provides a vital step forward in understanding the exciting world of AGN and their impact on galactic evolution!