Research > Outflows

Some of my results on outflows driven by supermassive black holes

My Related Popular/Press Articles:

“The Quasar Feedback Survey: Multi-phase outflows, turbulence and evidence for feedback caused by low power radio jets inclined into the galaxy disk”

Figure from Girdhar et al. 2022, MNRAS, 512, 1608, (2022), arXiv:2201.02208, ADS

In this study we present multi-wavelength observations of one of the targets, J1316+1753, including MUSE, ALMA and VLA. The source has small (~1kpc) and low power (Pjet~10^44 erg/s) radio jets that are included into the host galaxy disk. Our data reveal that these jets are causing high levels of outflowing turbulent gas in both the ionised gas and the molecular gas. The ionised gas, traced via [O III], has velocity widths reaching ~1200 km/s, over (at least 7.5kpc). This gas is expanding in an apparent cone above and below the disk. The molecular gas, traced via CO(3-2), has smaller velocity widths of ~400 km/s and extended to about ~4.5kpc. The jets are seen to be strongly interacting with the interstellar medium (ISM) through enhanced ionised emission and disturbed/depleted molecular gas at the jet termini.

We see further evidence for jet-induced feedback through significantly higher stellar velocity-dispersion aligned, and co-spatial with, the jet axis (<5 deg). We discuss how jet-induced feedback could be an important feedback mechanism even in bolometrically luminous ‘radio-quiet’ quasars.

The Figure shows, a summary set of figures from Girdhar et al. (2022) for J1316+1753. The left-most panel is a schematic summary of the main findings of the study. The central panel is a stellar velocity dispersion map, not the enhanced stellar velocities along the jet axis. The right-upper panel is a velocity width map (W80) for the molecular gas (CO(3-2)), note the enhanced velocity dispersion expanding above the jet. The right-bottom panel is a velocity width map (W80) for the ionised gas ([O III]), note the cones of ~1000 km/s gas both above and below the jet axis.

Girdhar et al. 2022

Extreme ionised outflows are more prevalent when the radio emission is compact in AGN host galaxies

Figure from Molyneux et al. 2019, A&A, 631, 132, (2019), arXiv:1909.05260, ADS

In Molyneux et al. 2019, we investigated the connection between radio emission and ionised outflows. We used the ~3000 SDSS-selected z<0.2 AGN from Mullaney+13 which have a radio detection in FIRST and/or NVSS radio surveys. We then determined which sources were extended or compact in the radio using a mixture of direct radio size measurements and morphological classifications and searched for ionised outflows using the [O III] emission line.

The Figure shows, example 1.4 GHz images (≈5 arcsec resolution from FIRST) and, where available, overlaid green contours from ≈1 arcsec resolution 1.4 GHz images (from our follow up VLA observations on a subset of targets). The insets show ≈0.3 arcsec resolution 6 GHz images (with blue contours). Synthesised beams are represented by appropriately coloured ellipses. Compact radio sources (top row) have their radio emission dominated within the SDSS fibre extent (magenta circles), whilst extended sources (bottom row) show significant emission outside of this. Right: Cumulative fraction of AGN with [O III] emission-line components with FWHMB greater than a given value, for: all sources (blue curve), compact radio sources (orange curve) and extended radio sources (green curve). Extreme ionised gas velocities are more prevalent for compact radio sources. This reveals a strong connection between radio emission and ionised outflows; possibly due to compact radio jets which are interacting with the host galaxy.

Molyneux et al. 2019

AGN outflows and feedback 20 years on

Figure from Harrison et al. 2018, NatAs, 2, 198, (2018), arXiv:1802.10306, ADS



In 2018 I wrote a review covering the challenges of measuring AGN outflow properties and understanding their impact (in particular in relation to simulations).

The figure shows a compilation of observationally determined kinetic coupling efficiencies for warm ionised gas from optical or near-infrared (NIR) emission lines (circles) and cold molecular gas from CO emission lines (squares). Hollow symbols are for single galaxies and filled symbols are for galaxy samples. The vertical lines show the range of values quoted, or in the case of an error bar, the quoted error on the average value. The horizontal lines show the range of bolometric luminosity for each sample. In the left panel we show the efficiencies that we found explicitly quoted as the representative/typical values in the abstracts (for warm ionised outflows) or within the papers (for cold molecular outflows) from our ADS search of the “active & galactic & nuclei & outflows” subset described in Figure 1 (see Table 1). The middle panel shows the compilation from Fiore et al. where kinetic powers were derived from literature observations using a uniform set of assumptions (no errors are provided by the authors). The horizontal dashed line in these two panels shows the “representative” (but variable) kinetic coupling efficiency calculated by Costa et al. The right panel shows coupling efficiencies employed by fiducial AGN feedback models, using the “quasar mode” value where applicable. In these models, a 100% of the injected energy is unlikely to become kinetic power in the outflow, so E_kin/L_AGN < εf , and direct comparisons to observed kinetic coupling efficiencies should not be made.

Harrison et al. 2018

Storm in a Teacup: X-ray, radio and ionised gas bubbles from a `radio quiet' quasar

Figure from Lansbury et al. 2018, ApJL, 856, 1, arXiv:1803.00009

Chandra Popular Releases about article


Using Chandra time (that I obtained as PI) my team observed the Teacup AGN, which we had previously identified as having large bubbles of ionised gas (green colour in image) and radio emission (red colour in image) to obtain spatially-resolved X-ray measurements. Using these data, in combination with existing XMM-Newton X-ray observations , we obtained the best constraints to date on the bolometric output (and level of obscuring material) of the growing supermassive black hole at the centre of this galaxy. These results revealed that the quasar may not be "fading away" as previously thought, based on earlier less robust constraints on the bolometric output. Furthermore, in the Chandra data we detected the large bubble in X-ray emission (see blue contours in image), providing tentative evidence for a very hot outflowing gas component within the bubble. This source appears to be showing outflows in multiple phases of gas, driven by the supermassive black hole lurking at its centre.

Lansbury et al. 2018

The prevalence of ionised outflows in high-redshift AGN

Figure from Harrison et al. 2016b, MNRAS, 456, 1195, (2016), arXiv:1511.00008

ESO messenger articles on this survey:


Using KMOS we have built up the largest ever sample of distant AGN with integral field spectroscopic observations. This KASHz survey is providing crucial information on the prevalence and properties of ionised outflows in the host galaxies of AGN during the peak epoch of galaxy formation. This figure, from our initial paper of the survey, shows the cumulative fraction of sources with an ionised gas velocity above a certain velocity. Velocities >600 km/s are indicative of outflow-dominated ionised gas kinematics and this is seen in ~50% of distant (z~1.5) AGN - shown by the blue curve. For AGN of matched power, this outflow prevalence agrees with that seen in the local Universe (z<0.4; grey curves).

Harrison et al. 2016

Galaxy-wide outflows are common among quasars

Figure from Harrison et al. 2014a, MNRAS, 441, 3306 (2014), (arXiv:1403.3086).

Gemini press/popular articles on this paper:

Using GMOS’s integral field spectrograph, we traced the properties of the gas across the host galaxies of 16 z<0.2 type 2 (obscured) quasars. Critically, GMOS on Gemini South enabled us to trace the velocity of the gas over the full spatial extent of the quasars’ host galaxies. By using emission lines produced by ionized hydrogen and oxygen to trace the gas velocities the observations revealed that outflows, reaching velocities of >1000 kilometers/second, were found over the full spatial extent of the host galaxies of all of the quasars observed (e.g., see figure). Estimates of the masses involved in these outflows indicate that they are removing significant amounts of gas from the galaxies and their overall properties are in line with model predictions. A key aspect of this work is the quasars that were observed with GMOS were initially selected from a parent sample of several thousand objects with optical spectra. These observations can thus be placed in the context of the overall population and reveal properties of quasars generally. For example, these observations imply that at least ~50% of the most luminous quasars (those predicted by models to drive the feedback mechanisms) exhibit outflows of this type and at least 70% of these outflows are extended of galaxy scales.

Harrison et al. 2014a