A Modified 3D Biconical Outflow Model: Spatial Constraints on AGN-driven Outflows

TL;DR

This work extends the 3D biconical AGN outflow framework by adding a rotating disk component and seeing convolution, enabling robust spatial constraints on ionized outflows in local type 2 AGNs. Using arcsec-scale mock data and Monte Carlo realizations anchored to SDSS type 2 properties, the authors constrain the launching velocity $V_{max}$ and outer opening angle $\theta_{out}$, finding most systems have $V_{max} \lesssim 1000$ km s$^{-1}$ and $\theta_{out} \approx 30^\circ$–$40^\circ$, with 2–5% of strong outflows reaching $V_{max} \sim 1000$–$1500$ km s$^{-1}$. The study also reveals that, under seeing-limited observations, outflow sizes inferred from velocity widths can be biased high when the angular size is comparable to or smaller than the seeing, yet the results remain consistent with a lack of global feedback in the local AGN population. Energetics estimates for ionized outflows suggest modest kinetic powers relative to $L_{bol}$ (roughly $10^{-6}$–$10^{-3} imes L_{bol}$), reinforcing the view that ionized outflows in nearby AGNs typically do not drive galaxy-scale quenching on their own. The framework provides a scalable, population-level approach to compare large samples, quantify outflow geometries, and interpret spatially resolved kinematics, while acknowledging limitations such as analytic flux/velocity prescriptions and the neglect of clumpy substructure.

Abstract

We present a modified outflow model and its application to constrain ionized outflow properties of active galactic nuclei (AGNs). By adding a rotating disk component to the biconical outflow model of Bae & Woo, we find that models with a rotating disk require faster launching velocities ($\lesssim$ 1500 km s$^{-1}$) than outflow-only models to be consistent with the observed gas kinematics of local type 2 AGNs. We perform Monte Carlo simulations to reproduce the observed distribution of gas kinematics of a large sample ($\sim$ 39,000), constraining the launching velocity and opening angle. While the launching velocity is moderate for the majority of the local AGNs, the notable cases of 2 - 5 % show strong outflows with $V_{max} \sim 1000-1500$ km s$^{-1}$. By examining the seeing effect based on the mock integral field unit data, we find that the outflow sizes measured based on velocity widths tend to be overestimated when the angular size of the outflow is comparable to or smaller than the seeing. This result highlights the need for more careful treatments of the seeing effect in the outflow size measurement, yet it still supports the lack of global feedback by gas outflows for local AGNs.

Figures (15)

• Figure 1: A schematic diagram of a biconical model example of Bae16. SMBH is located at the center of this plot, and the dust plane is shown as a red line between two cones. The dark gray area represents where the outflowing gas exists (i.e., between inner and outer opening angles), while the light gray area shows a hollow region. The LOS direction is shown as a black arrow on the right.
• Figure 2: 2D map examples of the modified model. From left to right, three 2D maps in each row present flux, velocity, and velocity dispersion of the corresponding model. First row: Bae16 biconical model. This model has the same geometry as Figure \ref{['model_plots']}. Second row: Disk model from GalPak$^\mathrm{3D}$. Third row: The bicone + disk combined model. Last row: The combined model after seeing convolution. Red dashed circles at the center and red bars on the top right show the FWHM seeing size. We also plot 3-aperture size as gray circles, which is the fiber size of SDSS spectra.
• Figure 3: Histograms of physical properties of Woo16 type 2 AGNs. From left to right, each histogram shows redshift, stellar mass, flux ratio between disk and outflow components, and extinction factor based on Balmer decrement.
• Figure 4: Model grids from the modified outflow model. Each color indicates a different launching velocity, 500 km s$^{-1}$ (cyan), 1000 km s$^{-1}$ (yellow), and 2000 km s$^{-1}$ (red): (a) The observed [O3] VVD distribution of type 2 AGNs Woo16. This color map is also plotted as contours and gray color maps in the other three panels; (b) Model grids with (solid lines) and without (dashed lines) the disk. Each line shows the change in kinematics depending on extinction at a fixed inclination; (c) Model grids showing the effects of bicone inclinations and extinction rates when the disk component is included. Solid lines indicate variations with extinction at a fixed bicone inclination, while dotted lines indicate variations with inclination at a fixed extinction rate; (d) Model grids with different opening angles and extinction rates. Each solid and dotted line indicates the same bicone opening angle and extinction rate, respectively.
• Figure 5: Left: Kullback-Leibler divergence values depending on the PDF of the launching velocity. The PDF of the opening angle is fixed to the best. The mean and standard deviation of $V_{max}$ first determine the Gaussian distribution, and we use the truncated Gaussian from 100 km s$^{-1}$ as a PDF of the $V_{max}$. The minimum Kullback-Leibler divergence is found when the mean and standard deviation are 250 and 400 km s$^{-1}$. Right: The same plot but $D_{max}$ from the 2D KS test. The minimum value is found when the mean and standard deviation are 350 and 300 km s$^{-1}$.