An Analysis of AGN-Driven Outflows in the Seyfert 1 Galaxy NGC 3227

Julia Falcone, D. Michael Crenshaw, Travis C. Fischer, Beena Meena, Mitchell Revalski, Maura Kathleen Shea, Rogemar A. Riffel, Zo Chapman, Nicolas Ferree, Jacob Tutterow, and Madeline Davis

ApJ: 2024, Vol. 971, Issue 17

 

Summary

 

What are we studying?

The process of accretion to an AGN can result in the expulsion of large amounts of radiation, which can push on the surrounding gas and drive it outwards. This drives a complex series of processes and interactions between the SMBH and its host galaxy that we broadly refer to as AGN feedback. This paper is a study of the AGN feedback processes in NGC 3227, which is a Seyfert 1 galaxy located at a distance of 23.7 Mpc. It has a dwarf elliptical companion, NGC 3226, with whom it’s tidally interacting.

 

This study focuses on characterizing the high-velocity outflowing gas that is frequently observed in feedback processes. By doing so, we also aim to develop an updated model for the biconical geometry along which the outflows travel, which will be essential towards better understanding how the feedback process operates in this galaxy.

Image credit: HST. This image shows NGC 3227 (left) extending its arm towards its faithful companion NGC 3226 (right).

 

How did we do it?

This figure shows contours of [O III] emission, which represent the narrow line region (NLR) outflows, taken from HST's WFC3. The green, blue, and pink slits represent the slit placements of the various instruments discussed in this study. We aim to study the NLR outflows, so many of the slits are intentionally positioned on top of the [O III] emission contours.

This study makes use of the following instruments:

  • Hubble Space Telescope (HST)'s Space Telescope Imaging Spectrograph (STIS). We used archival G750M (optical) observations in the immediate vicinity of the SMBH, which allows for extremely detailed analysis. We also use HST's Wide Field Camera 3 (WFC3) for [O III] photometry of the nucleus.
  • Apache Point Observatory (APO)’s Kitt Peak Ohio State Multi-Object Spectrograph (KOSMOS). We took measurements in optical wavelengths across a few position angles of NGC 3227, and because these slits are much larger than the STIS slits, we could observe large-scale patterns in the kinematics.
  • Gemini North Near-Infrared Field Spectrograph (NIFS). We used archival measurements of the J-, K-, and Z-bands to analyze the motions of near-IR gas within the innermost 3” x 3” of the AGN.
  • Atacama Large Millimeter Array (ALMA). We used data generously supplied by Alonso-Herrero et al. (2019) to analyze the kinematics in radio wavelengths.

We use a Bayesian algorithm called BEAT (Fischer et al. 2017) to fit multiple Gaussian profiles to one-dimensional spectra extracted from the two-dimensional data. We were able to fit both the narrow and broad emission lines, which increased the accuracy of our fits. An example of our fits are shown below. On the left, we fit Hβ; and the [O III] 5007/4959 doublet; on the right, we fit Hα and the [N II] 6583/6548 doublet. Fitting these lines will provide us with crucial information about how the gas is moving (for example, whether it's redshifted or blueshifted relative to the rest frame of the galaxy), and how bright the region is.

Notice above how this fit includes a mix of broad (olive) and narrow (teal, red, and orange) curves. If we take the plot on the left, we see that Hβ is fit with a broad component, but all three emission lines are each fit with two narrow components (that's the red and teal curves that you see under Hβ and both [O III] lines). Each narrow-line component is representative of a cloud or knot of gas moving at a distinct velocity, so the observation of two narrow-line components for this spectrum implies that at this specific location, the spectrograph has picked up the existence of two knots.

What did we find?

Successfully implementing BEAT to fit our data yields kinematic plots such as the one seen below, which reveals the Hα velocities for a KOSMOS slit. The red, blue, and green colors are representative of the first, second, and third Gaussian components that were fit to the data. Not all fits have second or third components, based on the output from BEAT. The black curve shows the rotation curve for NGC 3227 (Riffel et al. 2017), so significant deviations from this curve, such as the data extending to -500 km/s at 5” (~600 pc), show evidence of outflows.

We also produce plots showing the Gaussian line widths and integrated flux values as a function of distance, shown in the middle and bottom rows of the figure above, which help us futher analyze the gas motions. Similar kinematic analyses were performed on the rest of the spectra to better understand how gasses at these various phases move in relation to each other.

We also use these kinematics to better define the orientation of the bicone, which constrains the direction of the outflowing gas away from the AGN. With input parameters for the bicone like inclination and half-opening angle, we can model the bicone and estimate the range of velocities we would expect to see at a given distance. We assume a linear acceleration and deceleration for the radiated gas, and define the model’s turnover radius as the distance where the latter overtakes the former, and is visually represented in the models as the distances where the velocities peak.


We compared the data to over 100,000 possible models for each STIS slit, and developed an algorithm to determine which model produced the optimal fit with the data. The above figure shows the comparison between a given model (the gray regions) and our data (the blue points). The rotation curve is shown as the black outline.

We arrived at the bicone model shown above. There is one side of the cone pointed towards us (shown in blue), and the other side of the cone is pointed away (shown in red) The bicone is bisected by the disk of the galaxy, represented by the gray oval. Any emission under the galaxy's disk is osbcured by the dust. The [O III] contours showing the NLR outflows are overlaid on top.

Radiative Driving

We assume that two processes govern the motion of gas away from the AGN: radiative pressure from the enormous amounts of energy pushing outwards in the form of AGN-induced radiative acceleration; and gravitational deceleration from the enclosed mass in the host galaxy (Proga et al. 2000; Ramírez & Tombesi 2012; Meena et al. 2021). Under this formalism, we utilize a relationship that charts the trajectories of gas as a function of the distance from the SMBH from where the gas was launched.

Looking at the graph below, we can see that gas which is launched from within the innermost few parsecs of the SMBH travel farther and faster than gas launched ffrom distances of 10 or 20 pc. Therefore, using our kinematic data from KOSMOS, we can deduce launch distances for this gas: these knots, at current distances of 30 – 600 pc (marked by the red dots), can be tracked back to launch radii of 2 - 20 pc. In the graph below, the turnover radius where deceleration overtakes acceleration occurs at 31 pc. We compare this observational turnover radius to the one we calculate from our bicone model, which has been previously studied in Meena et al. (2023).

We find that the observed and modeled turnover radii of NGC 3227 are strongly correlated, implying that radiative driving alone is sufficient to replicate the kinematics that we observe.

For more details, be sure to check out the paper!