2406 03218v1

Draft version June 6, 2024
Typeset using LATEX default style in AASTeX631
A case study of gas impacted by black-hole jets with the JWST:

outflows, bow shocks, and high excitation of the gas in the galaxy IC5063
Kalliopi M. Dasyra,1 Georgios F. Paraschos,2 Francoise Combes,3 Polychronis Patapis,4 George Helou,5
Michalis Papachristou,1 Juan-Antonio Fernandez-Ontiveros,6 Thomas G. Bisbas,7 Luigi Spinoglio,8
Lee Armus,9 and Matthew Malkan10
1 Department of Astrophysics, Astronomy & Mechanics, Faculty of Physics, National and Kapodistrian University of Athens,
Panepistimioupolis Zografou, 15784 Athens, Greece
arXiv:2406.03218v1 [astro-ph.GA] 5 Jun 2024
2 Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, D-53121 Bonn, Germany
3 Observatoire de Paris, LERMA, Collège de France, CNRS, PSL University, Sorbonne University, Paris, France
4 Institute of Particle Physics and Astrophysics, ETH Zurich, Wolfgang-Pauli-Str 27, 8093, Zurich, Switzerland
5 Caltech/IPAC, 1200 E. California Boulevard, Pasadena, CA 91125, USA
6 Centro de Estudios de Fı́sica del Cosmos de Aragón (CEFCA), Plaza San Juan 1, E-44001, Teruel, Spain
7 Research Center for Astronomical Computing, Zhejiang Lab, Hangzhou, 311100, China
8 Istituto di Astrofisica e Planetologia Spaziali (INAF–IAPS), Via Fosso del Cavaliere 100, I-00133 Roma, Italy
9 California Institute of Technology
10 Department of Physics and Astronomy, UCLA, Los Angeles, CA 90095-1547, USA
ABSTRACT
We present James Webb Space Telescope MIRI data of the inner ∼3×2kpc2 of the galaxy IC5063,
in which the jets of a supermassive black hole interact with the gaseous disk they are crossing. Jet-
driven outflows were known to be initiated along or near the jet path, and the stability conditions of
clouds were known to vary because of these outflows. The MIRI data, of unprecedented resolution
and sensitivity in the infrared, now reveal that there are more than ten discrete regions with outflows,
nearly doubling the number of such known regions. Outflows exist near the radio lobes, at the nucleus,
in a biconical structure perpendicular to the jet, and in a bubble moving against the disk. In some
of them, velocities above escape velocity are observed. Stratification is also observed, with higher
ionization or excitation gas attaining higher velocities. More outflows and bow shocks, found further
away from the nucleus than the radio lobes, in regions without significant radio emission, reveal the
existence of past or weak radio jets that interacted with the interstellar medium. The coincidence
of the bow shocks with the optical extended emission line region (EELR) suggests that the jets also
contributed to the gas ionization. Maps of the H2 gas excitation temperature, Tex , indicate that the
molecular gas is most excited in regions with radio emission. There, Tex is >100 K higher than in the
EELR interior. We argue that a combination of jet-related shocks and cosmic rays is likely responsible
for this excess molecular gas excitation.
Keywords: Interstellar medium — Molecular gas — Galaxy winds — Radio jets — Galaxies
1. INTRODUCTION
The advent of the James Webb Space Telescope (JWST), with its unprecedented spatial resolution and sensitivity
in the infrared, enabled the astronomical community to make fast advancements in the field of galaxy evolution, by
looking further in time for distant galaxies and deeper into nearby galaxies. Besides the detection of the first galaxies
and even the first stars in them (e.g., Atek et al. 2022; Castellano et al. 2022; Naidu et al. 2022; Finkelstein et al.
2022, 2023; Donnan et al. 2023a,b; Harikane et al. 2023a; Leung et al. 2023; Maiolino et al. 2023a; Yan et al. 2023),
a major galaxy evolution theme to be revised by the JWST is the impact of black holes (BHs) on galaxy evolution
(including via outflows; Armus et al. 2022; Rupke et al. 2023; Vayner et al. 2024). Cycle 1 surveys already indicated
that many more active galactic nuclei (AGN) exist early, above redshift z>3-4 (Barro et al. 2023; Juodzbalis et al.
2023; Kocevski et al. 2023; Labbe et al. 2023; Lyu et al. 2023), compared to what was known from past observations
2 Dasyra, K. M., et al.
in X-rays or other wavelengths, and that some black holes grew too fast too soon compared to the local BH mass vs.
host galaxy relation (e.g., Harikane et al. 2023b; Larson et al. 2023; Maiolino et al. 2023b for 4<z<8.5).
If black holes started accumulating their mass earlier, it is likely that their radio-mode feedback phases, operating
via radio jets at low Eddington accretion rates, could have started earlier. The radio-mode feedback, statistically
operating after the radiation-mode feedback in simulations, has long been deemed critical for the reproduction of the
observed number of massive galaxies in the local Universe (e.g., Bower et al. 2006; Croton et al. 2006). Therefore, BH
jets could have propagated earlier, through younger galaxies with more massive interstellar medium (ISM) reservoirs.
Follow up of radio catalogues with the JWST and the Atacama Large Millimeter Array (ALMA) indeed led to the
discovery of new radio galaxies in the early Universe. For example, in the COSMOS field alone, Endsley et al. (2022)
and Labrides et al. (2024) detected radio galaxies at z∼6.8 and z∼7.7, respectively. Both of them have intense star
formation. Similarly, one of the major results of a recent ALMA survey of radio galaxies (Audibert et al. 2022) was
that the gas reservoirs of radio galaxies at 1<z<3, when detected, are comparable to the gas reservoirs of normal
galaxies at the same redshift. This result has important implications for galaxy evolution, because radio jets have a
tremendous potential to disturb star formation when propagating through gas-rich hosts. Hydrodynamical simulations
have shown that when relativistic jets pass through a gaseous clumpy disk, they contribute to the disk dissipation
by greatly dispersing dense clouds (depending on the ISM geometry, clumpiness, and density) and by forming a mass
loaded wind that entrains even more material (Wagner & Bicknell 2011; Wagner et al. 2013; Gaibler et al. 2012;
Mukherjee et al. 2018).
The turmoil that jets cause as they propagate through a dense ISM is demonstrated in this paper, with the aid of
cycle 1 JWST data of the nearby, early-type galaxy IC5063 (taken for the program 2004; PI Dasyra). IC5063 is well
suited for demonstrating the capabilities of the JWST, as it is a galaxy known for its numerous jet-cloud interactions.
These interactions occur as radio jets cross a disk of dense gas for hundreds of parsecs, along a trajectory nearly
parallel to the disk plane, prior to escaping it (Morganti et al. 1998). The jets collide with dense ISM clouds at the
two radio lobes, ∼300-500 pc away from the nucleus, driving outflows. An HI outflow has been seen in absorption
in front of one of the radio lobes (Oosterloo et al. 2000; Morganti et al. 2007), and in emission in several regions, in
tracers of all gas phases - atomic (Morganti et al. 2007; Dasyra et al. 2015; Congiu et al. 2017; Venturi et al. 2021)
and molecular (Tadhunter et al. 2014; Morganti et al. 2015; Dasyra et al. 2015, 2016, 2022). In Very Large Telescope
(VLT) SINFONI data, Dasyra et al. (2015) found numerous outflows starting at or near the jet trail: at least 6 discrete
outflow starting points were reported for the central 2×2kpc2 of this galaxy, including at the two radio lobes. In some
of these locations, gas moving faster than the escape velocity was detected. Moreover, the gas in nearly all of the inner
1.5 square kiloparsec of the galaxy shows distorted kinematic signatures due to the jet. In VLT MUSE data, Venturi
et al. (2021) showed that, in the ionized gas as seen by [O III], the turbulence in the galaxy may reach ∼4 kpc in both
directions perpendicular to the jet. Morganti et al. (2015) proposed even the propagation of a homonegeous jet-inflated
cavity (a.k.a. cocoon) in the inner region of IC5063 based on the low rotational number J CO data distribution. The
mass of all outflowing gas is at least 6×106 M⊙ , when taking into account the mass of the neutral gas and the cold
molecular gas, using their most-abundant gas tracers, HI and CO (Morganti et al. 2007; Dasyra et al. 2016). The mass
of the outflowing cold H2 required a CO luminosity to H2 mass factor different than Galactic (Dasyra et al. 2016; see
also Bolatto et al. 2013), because dissipated, optically thin molecular gas exists in regions with outflows (Dasyra et al.
2016; Oosterloo et al. 2017).
IC5063 has a large and rich set of available ancillary data, which allow novel experiments on the impact of jets on star
formation to be carried out. In a recent paper, Dasyra et al. (2022) showed that ongoing star formation (SF) changes
linked to the jet can be sought via stability analysis of jet-impacted clouds. The concept behind that paper was simple:
whether a cloud will collapse, and contribute to the enhancement of SF, or disperse, and add to the suppression of
SF, depends upon the interplay of its self gravity with changes in its internal and external pressure. The pressure
measurement can be challenging, yet it is feasible. It requires radiative transfer modeling of the line emission, if the
latter is assumed not to be in local thermodynamic equilibirum (LTE). For the cold and dense molecular gas clouds,
Dasyra et al. (2022) derived the internal pressure Pi by fitting the CO (1-0) up to (4-3) and HCO+ (4-3)/(1-0) ALMA
data with the astrochemical and radiative transfer code (3D-PDR; Bisbas et al. 2012), taking into account various
heating sources: UV radiation, jet-related cosmic rays, and mechanical heating (associated with shocks). They fitted
3D-PDR models to each pixel of the spatially resolved ALMA data and found that Pi increases from 5×104 K cm−3
(in ambient clouds a few hundred pc away from the jet trail) to 106 K cm−3 (in jet-impacted clouds at the radio lobes).
For the external pressure of the molecular clouds Pe , they considered it to be the pressure of the surrounding ionized
Jet-driven outflows, bow shocks, and high excitation of the gas in the MIRI data of IC5063 3
medium, which was derived from standard diagnostics of the ionized gas temperature ([N II]) and density ([S II]) in
MUSE data that are applicable for Boltzmann populations under LTE. Similarly, they found an increase of more than
one order of magnitude at the radio lobes for Pe . They also found that Pe frequently exceeds Pi , which led them
to conclude that we are observing the expansion of an overpressurized ionized gas bubble, perhaps the jet cocoon,
that engulfs and compresses clouds as it propagates. Simultaneously, in other, underpressurized regions, plugging Pi
and Pe in a modified virial equation that is appropriate for molecular clouds indicated that some clouds can become
gravitationally unbound and lose outer layers. This stability analysis therefore indicated that both star formation
induction and suppression can occur in IC5063 (Dasyra et al. 2022).
We obtained new JWST data with the goals of studying the molecular and ionized gas distribution, excitation, and
kinematics, finding the extent and mass of jet-driven outflows, comparing the loci of these outflows in unobscured
wavelengths with the loci of over/under-pressurized regions, and performing another calculation of the molecular gas
pressure at gas layers directly probed by the warm H2 . Many of these goals are addressed in this paper, whereas the
gas pressure will be addressed, with the use of detailed shock modeling, in a future paper.
2. THE DATA
We acquired data of IC5063 with the MIRI instrument onboard the JWST for the cycle 1 program 2004 (PI: Dasyra).
MIRI was used in its medium resolution spectrometer (MRS) integral-field-unit (IFU) mode (Wells et al. 2015; Argyriou
et al. 2023). The observations were executed the 13th of May 2023 and targeted the inner 13′′ ×8′′ of the galaxy in
ch1, and up to the inner 19′′ ×14′′ in ch4, in an area of the nucleus of IC5063 that follows the radio jet trail (Fig. 1).
A mosaic of four different tiles was used to cover the area of interest, because the MIRI IFU field of view (FOV), is
3.3′′ ×3.7′′ wide in channel 1 and 6.6′′ ×7.7′′ wide in channel 4. Additionally, a four-dither pattern was used for each
tile, to assist with drizzling and with the identification and elimination of problematic pixels. In each pointing position,
the exposure was either executed in a single up-the-ramp integration or split between two up-the-ramp integrations,
following the wavelength-dependent saturation indications of the Exposure Time Calculator (ETC). One up-the ramp
integration per exposure was used for the short detector (channel 1 and 2) data, and two up-the ramp integrations
were chosen for most long detector (channel 3 and 4) data. Separate sky observations for the background elimination
were also carried out. They had an identical exposure time and the same number of integrations per exposure to that
of the science observations, for a single mosaic tile and a four-dither pattern.
The level 1 data were reduced with version 1.12.5 of the JWST (Python-based) pipeline and the jwst 1202.pmap
content of the Calibration Reference Data System (CRDS). In the first stage of the pipeline, the function ”Detector1”
was run, which primarily corrects for detector artifacts and produces detection rates from time-dependent slopes. All
parameters were set to their default values for exposures with two integrations. Then, the values of pixels with cosmic
ray (CR) hits in one integration were replaced by their values in the other integration. For exposures with a single inte-
gration, ”Detector1” was run with the additional use of the ”find showers” parameter, so that CRs could be identified.
Their flux was modeled and subtracted from the relevant data by the pipeline. We performed two extra reduction steps
of our own to improve the data products. The first step was the subtraction of the background from the science frames.
A master background was created for each band (channel and sub-band) as the median of the four available exposures.
It was then subtracted from each science frame of the same band. This two-dimensional background subtraction
had the optimal result with respect to the removal of artifacts compared to the one-dimensional or three-dimensional
subtraction that is offered at later stages of the pipeline. Our second additional step was the masking of bad pixels in
the science data using a master bad pixel mask per band. We created this mask by merging the CRDS mask file with
a mask of outliers (>5σ) in the master sky frame. In the background-subtracted science data, the replacement values
of bad pixels were computed by interpolation of their neighbors’ values using the ”Gaussian2DKernel” function of the
”astropy” package and a standard deviation of one pixel.
We input the background-subtracted data to the second stage of the pipeline, ”Spec2Pipeline”, for all relevant cali-
brations prior to the building of the 3-dimensional cube. These include the world coordinate system (wcs) information
buildup, the flatfielding, the flux calibration, the straylight correction, the fringe removal, and the rectification of the
data. Besides all functions that run by default, we also enabled the residual fringe removal function (and disabled
the background subtraction function). During this step, we also created a version of the cubes without the straylight
correction, to check for the impact of the cruciform artifact (see Section A.1).
Calibrated cubes were created, which were then processed by ”Spec3Pipeline”, for the final stage of the pipeline
reduction, which merges individual exposure and pointing position cubes into mosaics of wcs-aligned cubes per band
Figure 1. IC5063 as seen in multiple optical bands, from data of two Hubble Space Telescope (HST) snapshot proposals
(15446 and 15444) and the Legacy Survey, DR9 release. Image credit: NASA/ESA/CSA/ J. Schmidt, attribution: NASA/ESA/
Aaron Barth / Julianne Dalcanton / DECaM Legacy Survey / Judy Schmidt, license: CC BY 2.0. For the Legacy Survey data,
DECam g is shown in blue, DECam r is in green and DECam z is in red. For the HST data, ACS/WFC F606W is in blue
and ACS/WFC F814W is in red (see also Maksym et al. 2020 for the HST data). In this image, north is 3.29° clockwise from
up. The inset shows the mosaic coverage in the four channels of the MIRI MRS observations for the cycle 1 program 2004 (PI
Dasyra). The contours show the radio emission (including the radio lobes) of the galaxy at 17.8 GHz (Morganti et al. 2007).
or channel. We opted to produce mosaics per band, for a custom handling of data in overlapping spectral regions.
The outlier detection and the default parameter values (e.g., drizzle for weighing) were used. Further 1-dimensional
residual fringe correction was applied. For the extraction of the spectrum of the nuclear point source, autocentering
of the aperture location was turned on. No further background subtraction was used, so all pertinent functions were
skipped (e.g., ”mrs imatch”). The footprint of the created mosaics, which increases with wavelength, is shown in
Fig. 1. Additionally, we converted the units of the final mosaic cubes from MJy/sr to Jy, taking into account the
channel-dependent pixel size.
The integrated spectrum of all the emission in the MIRI mosaic of each band is shown in Fig. 2. An impressive
number (39) of spectral lines is detected, most of which have spatially-resolved emission (see Section 3.1). Unlike the
Figure 2. Integrated spectrum in the entire mosaic of channel 1 through channel 4 (from top to bottom).
spectral lines, the mid-infrared continuum emission of IC5063 is nearly unresolved, originating in its largest fraction
from the nuclear point-spread function (PSF; as seen in A.3). Due to flatfielding uncertainties in the long band of
channel 4, its flux is on average lower (by ∼8%) than that indicated by the flux in previous channels. This calibration
uncertainty also applies to fluxes of the lines in the same channel and band. It is comparable to the nominal flux
calibration accuracy of the instrument (10%).
Continuum-free cubes have been created with own Python routines for each spectral line using a local continuum
image. For this purpose, a continuum image was first created from cube slices of emission-line-free wavelengths,
indicatively 3000-4000 km s−1 away from the spectral line rest-frame wavelength. Then, this image was scaled in
integrated flux to match the continuum flux per wavelength, according to a first degree polynomial that best fits
the continuum emission in the spatially integrated spectrum (Fig. 2). Because of the strength of the continuum, the
residual noise inside the nuclear PSF that appears after the continuum subtraction can be non-negligible compared to
the emission of several spectral lines. This means that the nuclear line fluxes are more uncertain that the line fluxes at
other mosaic locations. To visually enhance the appearance of the presented line cubes and images in this paper, we
created smoothed cubes, convolved by a Gaussian kernel of 1 pixel (for channels 1,2) or 1.5 pixels (for channels 3,4).
We then replaced the pixel values of the original-resolution cubes inside the PSF (only) with those of the smoothened
cubes.
The lines were found to have an average z=0.01121. For an adopted ΛCDM cosmology, with H0 =70 km s−1 Mpc−1 ,
ΩM =0.3 and ΩΛ =0.7, the galaxy has a luminosity distance of 48.4 Mpc. An arcsecond in the sky corresponds to 230
pc in the galaxy at that distance.
3. RESULTS
3.1. Overall emission properties and molecular disk properties
The continuum emission is largely unresolved, with more than 96% of its flux in all bands originating from the
supermassive black hole vicinity, at 20:52:02.36-57:04:07.54, according to the location of the PSF in ch1. Its flux is
0.22(±0.01), 0.95(±0.02), 1.28(±0.02), and 3.0(±0.3) Jy at 6, 12, 14, and 25 µm respectively. In the MIRI wavelengths,
we primarily see the AGN-heated dust, which is likely located in the black hole torus and which has been attributed
a temperature of ∼720 K for its near infrared emission (Kulkarni et al. 1998). The torus has been predicted in the
past from datasets of lower spatial resolution (e.g., Kulkarni et al. 1998; Hoenig et al. 2010; Esparza-Arredondo et al.
2019). It has been considered responsible for the detection of crepuscular rays in this galaxy (Maksym et al. 2020).
Its maximum extent is limited, from the PSF in our ch1 data, to ∼50 pc, which improves the same measurement
performed with ground-based mid-infrared sub-arcsecond observations (Asmus et al. 2014). A small contribution from
the radio core to the nuclear spectrum is also possible, given that some continuum emission (up to 4%) exists between
the nucleus and the radio lobes, which can be attributed to the jet-related synchrotron. Some weak silicate absorption
is also seen at 9.7 µm, with optical depth τ9.7 =0.18, which indicates the existence of compact dust in or around
the torus (and the radio core). The diffuse dust emission in the MIRI FOV is very weak compared to the nuclear
point-source emission. Still, the spatially-resolved PAH emission, in particular that of the brightest 11.3µm complex,
confirms its presence.
In contrast to the continuum emission, most of the line emission is spatially resolved. Thanks to the unprecedented
quality of the MIRI data, 39 lines of molecular and atomic gas phase were detected, and spectral cubes were created
for all of them except for the polycyclic aromaric hydrocarbons (PAHs). Images of the integrated line emission are
shown in Fig. 3 and their total fluxes are presented in Table 1. The comparison with the previous Spitzer data is
impressive. Less than half of the Table 1 spectral lines were previously present in spatially-integrated Spitzer Infra-Red
Spectrograph (IRS) spectra (Tommasin et al. 2010; Panuzzo et al. 2011; Rampazzo et al. 2013), even though the IRS
field of view was comparable to that of our MIRI mosaic for the short-high (SH) and long-high (LH) modes, and
greater than that of our MIRI mosaic for the short-low (SL) and long-low (LL) modes. A bare minimum of kinematic
information (i.e., a linewidth) was extracted for a handful of lines in the SH and LH modes (Guillard et al. 2012).
Detailed kinematical analysis is now possible for nearly all lines, revealing out-of-dynamical-equilibrium components
in many of them, e.g., in all of the previously detected or undetected H2 lines - from S(1) to S(8). Depending on the
line under examination, the emission probes the disk and regions within it or near it that are excited by the jet, or
even the nucleus.
The spectral line that best probes the disk extent and properties is H2 S(1) (Figs. 3; 4). Faint spiral arms can be
seen in all H2 data from S(1) to S(3). In the upper part of the disk, a spiral arm starting north-east of the nucleus and
Figure 3. Detected gas emission, indicatively grouped by gas phase, in units of 10−20 W m−2 . North is up. Crosses mark the
location of the nucleus and the EELR bases, NW and SE of the nucleus, as derived from the ionized gas emission at rest-frame
velocity in the MIRI data. Following Morganti et al. (1998), the EELR bases are considered to be the two radio lobes.
heading north-west of it, can be seen in all upper panels of Fig. 4. In the lower part of the disk, a fainter spiral structure
extending from south-west of the nucleus to south-east of it can be seen in Fig. 4 for molecular gas velocities in the
range 30<V<200 km s−1 range. Spiral structure with similar orientation has been seen before in CO interferometric
data (Dasyra et al. 2016, 2022). Given the detected disk rotation pattern that has been computed as the first moment
of the H2 S(1) cube (Fig. 4), then the upper side of the disk is closer to us if the spiral arms are trailing. The detected
dust lane location, in the north part of the galaxy (e.g., Fig. 1), confirms this and verifies that the spiral arms are
trailing. Several clumps bright in molecular gas, even brighter than those in the spiral arms, are seen along the radio
jets axis: in the line connecting the north west (NW) to the south east (SE) radio lobe, the brightest clumps are seen
Figure 3: Continued.
near the radio lobes themselves for all H2 lines, from S(1) to S(8). Still, other clumps also exist further away from the
nucleus, in particular for the lower rotational number lines. For example, in the south-east, two bright discrete clumps
are seen in H2 S(1) ∼ 900 and 1200 pc away from the nucleus, for positive disk velocities (identified as clumps b1 and
b2 in Fig. 5, where individual regions of interest are named).
In contrast to the molecular gas, the ionized gas lines do not show any spiral structure at disk velocities. The
emission of most these lines largely follows the jet direction from the nucleus all the way to the radio lobes, where
intense jet cloud-interactions are known to be taking place (e.g., Oosterloo et al. 2000; Morganti et al. 2007; Tadhunter
et al. 2014; Dasyra et al. 2015). The two radio lobes coincide with the bases of the two cones of the extended emission
line region (EELR) (Morganti et al. 1998), which makes the galaxy appear roughly X-shaped in its large-scale ionized
Figure 4. Properties of the disk, as best seen in H2 S(1). Upper and middle rows: Images of the emission at different velocity
ranges showing the rotation and structures of interest. Lower row: Moment 1 and 2 (velocity and velocity dispersion) maps.
gas distribution in optical wavelengths (Colina et al. 1991). In the MIRI data, the brightest locations of the ionized
gas emission are indeed the two the EELR bases (or radio lobes) and the nucleus. Lines of higher ionizational potential
(IP) tend to have more nucleated emission than lines with low IP. For all lines with IPs >100 eV (shown after [Fe VII]
in Fig. 3), more than two thirds of the emission is circumnuclear, linked to the AGN and to the radio core. From
the radio lobes outwards, the emission splits up in two main branches (see, e.g., [Mg V] and [Fe VII] in Fig. 3 for the
two branches north-west of the NW radio lobe and [S IV] and [Ne VI] for the branches south-east of the SE radio
lobe), which is why we alternatively call the EELR bases as bifurcation points. Further out, these branches constitute
the edges of the EELR, which is long known to be oriented at a position angle similar to that of the disk instead of
perpendicular to it (Colina et al. 1991). Just like in the optical, the EELR orientation is similar to that of the jet in
the mid-infrared. At distances greater than the bifurcation points, all of the ionized gas emission is constrained in the
interior of the EELR. Good tracers of the jet-ionized gas along the jet trail and along the EELR edges are [Mg V]
and [Ne VI], thanks to their high IPs (109.3 and 126.2 eV, respectively) and to the high resolution of MIRI at their
wavelengths. We further explain the connection of the gas emission with the jet in the sub-section that follows.
Figure 5. Structures of interest including bow shocks, filaments, discrete outflows, and structures moving against the disk at
positive velocities. Clouds, filaments, bow shocks/bubbles, and spiral arms are respectively marked with ‘c’,‘f’, ‘b’, and ‘sp’.
Figure 6. Structures of interest including bubbles, filaments, discrete outflows, and structures moving against the disk at
negative velocities. The nomenclature is as in the previous Figure.
Figure 7. Line profiles indicative of outflows. All spectra are extracted within 1′′ from the indicated location. Upper row
panels: Normalized spectra of the shocked gas tracers H2 and [Fe II]. Lower row: Normalized spectra of other ionized gas lines,
sorted by IP. Only spectral lines with signal-to-noise ratio >30 are plotted, so that line wings can be reliably compared.
3.2. Regions influenced by the jet and outflows

Bow-shock-shaped structures are revealed for the first time for this galaxy, already for velocities as low as those of
the gaseous disk, i.e., 260 km s−1 (as deduced from the H2 S(1) first moment map of Fig. 4; see also Fonseca-Faria et
al. 2023). A particularly prominent such structure is the H2 -bright region marked as b1 in Fig. 5, 1200 pc south-east
of the nucleus. Even though in the low-resolution H2 S(1) data it appears like a clump, in line with higher resolution
data, i.e. in H2 S(3), S(4), and S(5), its shape is very clearly indicative of a bow shock that is headed outwards in the
direction of the SE branch of the ionized gas EELR (marked as f3 in Fig. 5). This structure is indeed a bow shock,
not a projection effect of spatially unrelated clumps, as it maintains its shape in cubes in which there is no detectable
emission by any other structure in its vicinity (such as a weak spiral arm). Another bow shock is b2, with a similar
orientation according to H2 S(3), S(4), and S(5). In H2 S(7), a third bow-shock-shaped structure is also seen at the
SE lobe itself, at ∼150 km s−1 (although this structure may be part of the ionized gas bubble b3; see below). The
larger-scale structure connecting b1, b2, and b3, as best seen for H2 S(1) in the velocity range 100<V<400 km s−1 ,
does not follow the spiral patterns described above (marked as sp in Fig. 5), but the jet axis and the EELR edge
instead. These findings suggest that the jet has traveled in the EELR and driven bow shocks in regions b1 and b2. In
b1 and all along the f3 filament leading up to b1 (shown e.g., in the [Ne III] image at V>300 km s−1 ; Fig. 5), a weak
outflow is seen in many emission lines (Fig. 7). In the north-west part of the galaxy, both the edges (f5, f6, f8; Fig. 6)
and the interior of the ionized gas EELR (f7 and f9) are rather filamentary. One of these structures, f9, is in a nearly
straight line with the jet axis, as seen at 17.8 GHz, again suggesting that the jet has propagated in the EELR.
In the 17.8 GHz radio image presented by Morganti et al. (2007), bright radio emission exists mainly at the radio core,
lobes, and between them. No significant emission is detected near b1. In the other side of the galaxy, contours shown
close to the NW radio lobe are elongated in a direction comparable to that of the f5 branch outer part, as delineated by
[Mg V]. A similar elongation is also seen in contours of a 8 GHz image (Morganti et al. 1998). Therefore, it is possible
Figure 8. First and second order moment maps (velocity and velocity dispersion) of indicative spectral lines, showing regions
with distorted kinematics. Pixels with flux values in the noise were not taken into account for the creation of these maps.
that some weak radio emission exists, to be revealed with higher sensitivity observations. It is also possible that a jet
passed through the region with relativistic speed and that we now see gas pushed and excited by the slower-moving
outflows that it raised. As shown in studies of other nearby galaxies, it is possible to detect the passage of a past or
of a weak radio jet by the impact it leaves on the ISM: Aalto et al. (2016, 2020), Fernández-Ontiveros et al. (2020),
Pereira-Santaella et al. (2022), and Papachristou et al. (2023) demonstrated that outflows can reveal the existence
of jets, even if these jets are not seen in radio wavelengths. As we will show below for IC5063, outflows exist at the
nucleus, at the radio lobes and in many other regions.
Spectra of the outflows seen in these three regions, which are frequently studied in the literature, can also be found
in Fig. 7. Overall, very wide line profile wings are seen close to the radio lobes, and in particular near the NW lobe,
where the outflowing ionized gas reaches projected velocities of 1000 km s−1 for many ionized gas lines. The widths of
the lines with high IP (around or above 100 eV) are wider, showing more prominent outflow components than [Ar II]
with 15.8 eV. The molecular gas attains lower velocities than the ionized gas there, up to 750 km s−1 . Still, the higher
the H2 rotational number, the more blueshifted the emission is. At the SE radio lobe, a similar trend is observed for
the molecular gas, and a second blueshifted component associated with an outflow becomes brighter with increasing
IP of the ionized gas species. These results indicate stratification of the gas in outflows, meaning that the closer the
gas is to the power source, the higher its excitation/ionization and its velocity are (e.g., Armus et al. 2022). The
stratification of gas in outflows was discovered already in Spitzer Space Telescope data of local AGN with fast ionized
gas outflows (Dasyra et al. 2011; see, for example, IRAS05189-2524 and IRAS15001+1433). It is now found to also
be possible in regions within galaxies with jet-driven outflows.
An examination of the cube slices at different velocity ranges indicates that there are, in fact, several starting points
of outflows within 1′′ from each radio lobe. In the overall MIRI FOV, twice as many spatially discrete outflows as
those already known from the VLT SINFONI data are now revealed (Dasyra et al. 2015). In the western part of the
galaxy, the fastest outflow with positive velocities (of ∼ 1000 km s−1 ) is seen in the clump c2, which is about midway
between the nucleus and the NW radio lobe, and which lies at the center of an overpressurized region in the MUSE
data (Dasyra et al. 2022). The second highest velocities are found in the clump c1, which is 100(±10) pc away from
the bifurcation point at the EELR base. At negative velocities, of about -1000 km s−1 , yet another clump (marked as
c7 in Fig. 6) is seen in [S IV] 90 pc north east of the same bifurcation point, and south of a region with high ratio of
outflowing to regularly rotating gas seen in SINFONI [Fe II] 1.644 µm data (Dasyra et al. 2015). All these regions
could either be discrete starting points of outflows or fragmented clouds in the same outflow. Beyond the bifurcation
point, a filamentary outflow is detected along the NW EELR edge (see e.g., [Mg V] at -600<V<-300 km s−1 ). Similar
findings can be reported for the eastern part of the galaxy. The outflow along the f3 filament in the EELR SE edge
extends out to 800 pc from the SE bifurcation point. An individual clump can be identified in it (c4). North of it,
closer to the other EELR branch (f4), two more clumps that are moving at high velocities are identified (c5,c6). In
c6, which is in the EELR interior and close to the jet axis, the velocities reach 900 km s−1 for, e.g., [S IV]. Another
outflowing clump closer to the nucleus is c3, seen up to 500 km s−1 for the same line.
A large-scale region that contains gas that is outflowing, moving away from us, is found on the western side of
the galaxy. It begins from its own discrete starting point about mid-way between c2 and the nucleus. It extends
perpendicularly to the jet propagation axis, both to the north-east and to the south-west of it for at least 500 pc in
each direction. It is marked as f1 in Fig. 5, and it is seen in velocities in the 0<V<400 km s−1 range, with a peak at
220±30 km s−1 . Its emission is redshifted, while that of the disk is blueshifted in the same region. It is most prominent
in shock-tracing lines, i.e., in all H2 lines from S(3) to S(7) and in [Fe II] 5.34 µm (Fig. 5). It has several discrete
clumps identified in it. It can also be seen in a few ionized gas lines of low IP, such as [Ar II]. This is the structure
that we called biconical outflow in the SINFONI data, in which we first identified it (Dasyra et al. 2015) because
the fraction of intermediate-velocity to low-velocity gas in it opens up as a cone with increasing distance from the jet
axis. In the SINFONI data, the biconical outflow was seen again in shocked gas tracers, and more specifically, in the
ro-vibrational H2 (1-0) S(1) and S(3) lines and [Fe II] 1.644 µm. In the MIRI data, the biconical shape makes this
outflow easily distinguishable in the first moment maps of several lines, including those of H2 S(5), [Fe II], [Ar II], and
[Ne II] (Fig. 8). High velocity dispersion is associated with this outflow too, with the greater turbulence detected in
the southern part of the galaxy (Fig. 8). In optical MUSE data, Venturi et al. (2021) had found large-scale turbulence
of the ionized gas, probed by an effective width of the [O III] line, that extends up to 4 kpc away from the jet axis in
both directions. These might be different scale components of the same outflow, or similar orientation outflows with
neighboring starting points, generated as the jet collided with neighboring clouds and scattered gas perpendicular to
Figure 9. Summary, shown as a schematic representation, of all identified regions with outflows. The nomenclature follows
that in Figs. 5 and 6. The disk plane is in grey. The distances and number of clouds per region are indicative. Red and blue
arrows indicate outflows receding from the observer or approaching the observer, respectively.
its path. In the eastern part of the galaxy, a similar structure, but much weaker in emission is found. It is marked as
f2 in Fig. 5 and it is near regions with negative velocity components south-east of the nucleus in the [Ar II] velocity
map.
The moment maps also indicate the existence of another clump moving at opposite velocities compared to those of
the disk. It is a cloud 300 pc north of the NW EELR base, seen in both the first and the second moment maps of
[Ar II], [Ne II], and [Fe II]. It is also seen in the dispersion map of the 1.644 µm [Fe II] line observed with SINFONI.
We call it c8 in Fig. 9, where a summary of all jet-cloud interactions is schematically shown.
Finally, one more newly discovered and prominent structure, identified for the first time in the JWST data, is a
very bright and clearly delinated bubble (b3 in Fig. 6). It is located between the SE lobe and the nucleus, and it
can be seen for many ionized gas lines of different IPs, even for HI Pfund α. It is indicatively shown in Fig. 6 for
the bright [Fe II], [Ar III], and [Mg V] lines and for negative velocities down to -300 km s−1 , displaying the bulk of its
emission. For negative velocities, its material is approaching us while it is moving against the regular disk rotation.
Still, the bubble also contains some gas moving away from us, seen mostly in the 0<V<100 km s−1 range. Overall, it
is a structure with irregular kinematics moving at low projected velocities. This structure emits also in [Fe II] 1.644
µm in SINFONI data, where it appears more like a blob due to the lower resolution of those data. This bubble is likely
tracing a jet-driven or a supernova-driven outflow centered mid-way between the nucleus and the SE radio lobe (with
a few bright clumps in it). In the latter scenario, it would make IC5063 a great candidate for studying jet-induced
star formation, as this supernova could be related to jet-induced star formation.
In total, we see more than ten outflowing clouds and structures in our data. The exact number depends on whether
some clumps can be considered part of a common outflow (that started from the same point and that is headed in the
same directions) or not, as the medium that we see is very clumpy. A series of jet-cloud interactions and reformation
of clouds in outflows (Zubovas & King 2014) could be contributing to the clumpiness of the medium. We note that it
is possible that even more outflows exist in our data, but that the cruciform artifact (see Section A.1) prevents us from
identifying them with certainty. Such an example is the structure u1. Similarly, some low S/N regions seen around
the nuclear PSF in the -1150<V<-850 km s−1 [S IV] image could either be real or due to PSF residuals.
We summarize all unambiguously identified regions with outflows in our data in Fig. 9, using a schematic represen-
tation. In this representation, the jet travels in the ISM and drives both receding and approaching outflows in regions
Figure 10. The area in the nucleus of IC5063 with kinematically distorted gas, as seen by the JWST. Highly turbulent regions
are indicated with the aid of the velocity dispersion contours of two bright lines of Fig. 8: [S IV] and [Ne V]. Contours are
shown at 105, 120, 140, 160, 200, 250, 310 km s−1 for [S IV] and 105, 130, 160, 210, 290 km s−1 for [Ne V]. The biconical outflow
is shown with the aid of the velocity contours of [Ne II], for V=20 km s−1 . The outflows along the NW and SE EELR edges
are shown with contours of the [Mg V] -600<V<-300 km s−1 flux image and of the [Ne III] 300<V<500 km s−1 flux image,
with respective contour levels of 3.5×10−21 and 9×10−21 W m−2 . The b1 bow shock contours are created from the H2 S(4)
220<V<360 km s−1 image, for a flux of 1.8×10−21 W m−2 . For the three flux images, contours are created only for the relevant
extended outflows (i.e., for the outflows further away than the radio lobes from the nucleus that are not already comprised in
the high-σ regions). All contours are plotted over the optical image of the galaxy that is shown in Fig. 1, from NASA / ESA /
CSA / J. Schmidt.
of direct collision with dense clouds, where gas is dispersed in all directions. Outflows characterized by redshifted
emission only originate from clouds that are located behind the radio jet with respect to the observer and that are
pushed further away by the jet. Inversely, outflows characterized by blueshifted emission only originate from clouds
that are located in front of the radio jet with respect to the observer and that are pushed closer to the observer by the
jet. This explanation holds irrespective of which galactic hemisphere each outflow is found in.
A plot visualizing all regions with high gas turbulence can be found in Fig. 10. It shows how wide-spread the
turbulence in the center of IC5063 is. In total, the gas kinematics are disturbed in ∼2/3rds of the ch1 FOV. Moreover,
turbulent medium appears to be moving in projection towards a dusty structure seen in the optical image of Figs. 1
and 10 that runs nearly parallel to the jet axis, ∼1200 pc north of it. This dustry structure comprises clouds with
clear signs of either ram pressure or radiation pressure having acted on them. It is thus possible that these clouds
have been affected by a past episode of jet or AGN radiation feedback.
Another diagram showing how ubiquitous outflows are in our data is the position-velocity diagram along the major
axis of rotation, constructed for the bright [S IV] line (Fig. 11). Theoretical curves of the rotational and escape velocities
as a function of radius are also overplotted. The gravitational potential associated with the matter distribution in
IC5063, at the radii under examination, is assumed to be well approximated by a Hernquist potential (Mukherjee et
al. 2018), for which analytic formulas describe the radial profiles of the rotational velocity and the escape velocity
(Hernquist 1990). For the mass and the scale length of the stellar distribution needed for this calculation, we also
Figure 11. Left: Pseudo-slit position angle and width used for the position-velocity diagram creation, plotted over the [S IV]
moment 1 map. The position angle is identical to the major axis of rotation, as derived from the H2 S(1) rotation field. Right:
Position velocity diagram along the major axis of rotation. Pixel values below noise have been masked in this diagram. The
black solid line shows the rotational velocity in the Hernquist bulge that best characterizes this galaxy (see the text for more
details). The black dotted line shows for completion the same velocity because of the black hole potential. The positive and
negative escape velocities from the Hernquist bulge are also plotted as a function of radius, using white solid lines. An inclination
correction is applied to all theoretical velocities, for consistency with the observations.
adopted the parameter values of Mukherjee et al. (2018). These relied on observations of the central stellar velocity
dispersion (160 km s−1 ) and effective radius (21.5 ′′ ) from Kulkarni et al. (1998) and Morganti et al. (2015), and
indicated a stellar mass of 1.67 × 1011 M⊙ and a scale length of 2.8 kpc for the stellar bulge. Correcting for the
inclination of the galaxy, with an inclination angle of i = 74◦ (Morganti et al. 2015), we obtained the rotational velocity
and escape velocity curves in the position-velocity diagram (Fig. 11). The position angle along which the [S IV] flux
was collapsed was identical to that of the major axis of rotation of H2 S(1), 108◦ (Fig. 4). Unlike the molecular gas
traced by H2 S(1), which never reaches projected velocities greater than the escape velocity, the outflowing ionized
gas probed by [S IV] indeed travels faster than the escape velocity nearly all along the inner 600 parsec of the galaxy.
Only in the ∼100 central parsecs this claim cannot be reliably made due to PSF residuals and the to contribution of
the black hole of 2.8×108 M⊙ (Nicastro et al. 2003) to the gravitational potential. In the south-east of the galaxy, the
faint outflow seen in Fig. 7 for the b1 bow shock region is also continuous, appearing all along the EELR edge from
600 to 1500 pc from the center (see also Wagner & Appenzeller 1989 for a blue wing in [O III] 5007 Å at ∼600 pc).
However, the gas projected velocities there do not reach the corresponding escape velocities.
3.3. Density of the ionized gas

The density of the ionized gas can be determined in the MIRI wavelength range with two sets of lines of the same
species and ionization potential: the [Ar V] 13.1µm/7.9µm flux ratio and the [Ne V] 24.3µm/14.3 µm flux ratio. We
use the former ratio, which has sufficient spatial resolution for the creation of a map. The density range that [Ar V]
13.1µm/7.9µm can measure is shown in Fig. 12. Its values are computed with the code Pyneb (Luridiana et al. 2015)
for different temperatures and for line intensities that are reflecting electronic levels that are populated based on the
Boltzmann distribution, i.e., for gas under LTE. The derived density map (Fig. 12) was computed for an indicative
temperature of 10000K. This temperature was deemed appropriate given that flux ratios > 1.5 were found in several
pixels of our data, and also given that previous temperature calculations from [N II] lines in MUSE data were in the
range ∼7000-12000K (Dasyra et al. 2022). The density map shows discrete regions of overdense ionized gas, with
densities somewhat higher than 10000 atoms per cubic centimeter in locations where outflow starting points are seen,
including c1 and c2. For comparison, for the vicinity of c2, Holden et al. (2023) report densities of 6000 cm−3 based
on [Ar IV] in the optical. Similarly, in the NW EELR (regions f6, f7), in a region with outflows in our data and with
particular polarization in HST FOC data (potentially due to the outflows; Barnouin et al. 2023), the density seems
elevated. Overall, it is noteworthy that the [Ar V] diagnostic ratio, which is applicable at rather high density values
(>1000 cm−3 ; Fig. 12) and ionization potential (59.8 eV), works well for so many regions of the nucleus of IC5063.
This result could plausibly be indicative of compression of the ionized medium.
=
]. 1.5 =
=
1.0
] . /[
0.5
[
102 104 106 108

( )
Figure 12. Left: The [Ar V] line ratio as a function of electron density in LTE conditions for different temperatures. Right:
[Ar V]-based electron density map. Crosses are as in the previous figures. The dashed circle indicates the region where the
density values are uncertain due to PSF residuals. Densities are in units of cm−3 .
Figure 13. [Ne III]/ [Ne II] ratio, integrated over all velocities, as an excitation indicator. The circle and crosses are as in
Fig. 12.
4. DISCUSSION: THE EXCITATION MECHANISM(S) OF THE GAS

4.1. Ionized gas
In the previous Section, we demonstrated that the jet plays a major role in the observed gas properties in the center
of IC5063. A series of outflows along and perpendicular to the known jet axis from radio wavelengths cause great
turmoil in the gas kinematics. Bow shocks out to ∼1200 pc and filamentary outflows all along the ionized gas EELR
SE branch, as well as outflows out to ∼800 pc along the ionized gas EELR NW branch, suggest that the bifurcating
appearance of the EELR in both sides of the galaxy is due to jet scattering at the radio lobes. Our findings corroborate
that the jet is far more important than the AGN radiation in exciting the gas than previously thought (Morganti et
al. 2007; Mingozzi et al. 2019; Venturi et al. 2021). Of course, the AGN radiation may still be contributing to the
gas excitation in regions previously cleared out by the jet, as indicated by the existence of AGN radiation and jet
inflated cavities in X-ray data (Travascio et al. 2011) and the existence of circumnuclear high-ionization line emission
in our MIRI data. Still, the jet contribution to the gas kinematics and excitation is now unambiguously proven even
further away than the radio lobes. Already since the presentation of the EELR of IC5063, Colina et al. (1991) pointed
out that the radial profile of the ionized gas emission does not agree with the radial profile of the ionized gas density
illuminated by a nuclear point source (i.e., the AGN). Instead, these two radial profiles require the existence of an
extended ionization source, i.e., of a radio jet, in order to agree. Given that in the eastern part of IC5063, the SE
branch is the one with the highest luminosity and most extended outflow of the two EELR edges, and that, in the
Table 1. Line properties
Line name λ f(a ) σ

(-) (µm) (10−17 W/m2 ) ( km s−1 )
H2 S(8) 5.053 1.28 (± 0.32) 130 (± 12)
[FeII] 5.340 5.64 (± 0.62) 176 (± 13)
[FeVIII] 5.447 3.17 (± 0.11) 84 (± 9)
[MgVII] 5.503 3.52 (± 0.56) 75 (± 8)
H2 S(7) 5.511 3.84 (± 0.52) 150 (± 30)
[MgV] 5.610 5.08 (± 0.48) 90 (± 13)
[KIV] 5.982 0.52 (± 0.10) 135 (± 15)
H2 S(6) 6.110 1.68 (± 0.29) 113 (± 20)
[CaVII] 6.154 0.46 (± 0.11) 71 (± 11)
PAH 6.2 6.220 11.87 (± 0.84) ...
[SiVII] 6.492 1.36 (± 0.35) 189 (± 31)
[NiII] 6.636 1.33 (± 0.29) 243 (± 39)
H2 S(5) 6.910 9.26 (± 0.87) 129 (± 13)
[ArII] 6.985 9.44 (± 1.80) 119 (± 19)
[NaIII] 7.318 1.26 (± 0.18) 138 (± 13)
HI Pfund α 7.460 1.93 (± 0.30) 144 (± 23)
[NiI]+PAH 7.7(b ) ∼7.5 1.10 (± 0.30) ...
[NeVI] 7.652 33.83 (± 1.27) 80 (± 8)
[FeVII] 7.815 1.44 (± 0.24) 96 (± 9)
[ArV] 7.902 2.60 (± 0.44) 188 (± 15)
H2 S(4) 8.025 4.60 (± 0.66) 128 (± 5)
[ArIII] 8.991 11.83 (± 0.99) 112 (± 13)
[NaIV] 9.041 0.63 (± 0.09) 250 (± 60)
[FeVII] 9.527 3.25 (± 0.38) 174 (± 35)
H2 S(3) 9.665 13.34 (± 1.40) 130 (± 7)
[SIV] 10.511 54.13 (± 3.70) 121 (± 13)
PAH 11.3 11.3 30.77 (± 6.05) ...
H2 S(2) 12.279 5.67 (± 0.99) 122 (± 16)
HI Humphrey α 12.370 0.77 (± 0.13) 162 (± 21)
[NeII] 12.814 28.53 (± 3.03) 126 (± 12)
[ArV] 13.102 2.41 (± 0.61) 89 (± 5)
[NeV] 14.322 29.75 (± 2.60) 90 (± 10)
[NeIII] 15.555 79.29 (± 5.30) 125 (± 14)
H2 S(1) 17.035 12.86 (± 1.16) 125 (± 17)
[FeII] 17.936 1.73 (± 0.45) 114 (± 24)
[SIII] 18.713 31.29 (± 1.37) 144 (± 14)
[NeV] 24.318 23.21 (± 2.31) 97 (± 8)
[OIV] 25.890 107.83 (± 6.51) 130 (± 10)
[FeII] 25.990 4.33 (± 0.75) 113 (± 33)
H2 S(0) 28.219 <3.6(c ) ...
(a ) Fluxes integrated over the entire FOV. The FOV differs from channel to channel, as shown in Fig. 1, but this has a small effect for the
integrated flux of most lines, as the bulk of the emission lies within the common FOV (Fig. 3). The error bar associated with each flux
corresponds to the difference between the value derived from the direct integration of the signal on top of the continuum in the spectrum
shown in Fig. 2 and the value derived when spatially integrating the pixel fluxes in collapsed images of continuum-free cubes (Fig. 3).
This way, each error bar encapsulates uncertainties in the continuum subtraction (and therefore exceeds the noise accumulated over the
line profile. (b ) These lines cannot be unambiguously distinguished. (c ) This line is undetected in all spectral pixels of the cube and in
their collapsed image. Given the measured redshift of this galaxy, the H2 S(0) line is anticipated in the very last ∼25 spectral pixels of
the ch4 long cube. The instrumental throughput was low there, dropping at the time of our observations, which led to a highly uncertain
flux calibration. A tentatively measured flux limit of 3.6×10−17 Wm−2 is below the flux provided in Guillard et al. (2012) from Spitzer
Space Telescope data. Therefore, this line is excluded from all our analyses.
western part of IC5063, the same holds for the NW branch, it is possible that the jet propagated at a higher position
angle than that of the disk. This scenario is also supported by the ionized gas excitation, if examined with the aid of
the flux ratio of two bright lines, [Ne III] and [Ne II] (shown in Fig. 13). A similar result was seen in the optical for
the [O III]/Hbeta flux ratio (Fonseca-Faria et al. 2023).
4.2. The molecular gas excitation

The warm molecular gas, as seen directly in the H2 rovibrational lines, has been known since Spitzer observations
to show enhanced emission in radio galaxies compared to star forming galaxies thanks to the presence of jet-related
turbulence (Ogle et al. 2010; Guillard et al. 2012). This finding is also excellently demonstrated in our new JWST data
of IC5063, in a spatially resolved manner. The excitation temperature of the warm H2 is shown in Fig. 14 for indicative
line ratios. The computation is made under the assumptions that the line emission originates from gas that is optically
thin and in LTE, so that all molecules are probed by photons and so that the Boltzmann distribution applies to the
rotational level populations. These are meaningful and widely-applied assumptions for the low rotational levels of the
MIR-emitting gas (e.g., Rigopoulou et al. 2002; Higdon et al. 2006; Roussel et al. 2007; Ogle et al. 2010; Dasyra &
Combes 2011; Guillard et al. 2012; Pereira-Santaella et al. 2022). The flux of two lines is needed for the computation of
the excitation temperature Tex , which can then be used for the mass computation. Tex can be computed by equating
the total number molecules as given by two lines, or equivalently, by calculating the negative of the inverse of the
slope between two points in the excitation diagram (Rigopoulou et al. 2002). The excitation diagram is shown as the
natural logarithm of the column density of electrons NJ that descended from an upper to a lower state, divided by
the statistical weight gJ of the transition, as a function of the temperature or the energy of the upper state. The
number of electrons that made the transition is computed as L/(αhν), where h is the Planck constant, α is the Einstein
coefficient of the transition, and ν is the frequency of the emitted line. Multiplied by the pixel size, this number gives
the desired column density.
The Tex map for the coolest gas in our data, that emitting in S(1) and S(2), is in many regions quite similar to the
same map computed for the ortho lines S(1) and S(3), which are brighter. This is an indication that the low-J lines
are thermalized in many regions of our FOV. The same conclusion can be drawn when examining the full excitation
diagram of some indicative regions (Fig. 14; lower panel). This result is meaningful in the context of the gas densities
measured so far: when roughly assuming that the H2 gas density is intermediate to that of [Ar V] and that of CO
(Dasyra et al. 2022), then it is of order 104 cm−3 . At such densities, the H2 rotational levels can be rather safely
considered thermalized up to at least J=5 (Le Bourlot et al. 1999).
The Tex maps based on H2 S(2) and S(1), S(3) and S(1), as well as S(4) and S(2) show some more intriguing results:
that the highest Tex values are seen at the vicinity of the radio lobes and that the temperature is lower by at least
∼100 K in the bow shock regions than in the region where synchrotron emission is detected, within ∼500 pc from the
nucleus (Morganti et al. 1998, 2007). Interestingly, this inner, synchrotron-emitting region also shows a step in the
cold molecular gas heating rate and excitation temperature in such maps derived from CO and HCO+ observations
(Dasyra et al. 2022). The CO density also happens to be elevated in that region, just like the ionized gas density in
the optical MUSE data (Dasyra et al. 2022). All these findings imply that there is an additional excitation mechanism
of the molecular gas that is linked to the jet.
Two jet-related mechanisms are highly likely candidates: either a superposition of shocks or cosmic rays (see also
Ferland et al. 2008; Nesvadba et al. 2010; Padovani et al. 2022; Leftley et al. 2024). Indeed, either mechanical heating
or cosmic ray heating or a combination of the two was deemed possible in our CO modeling with the 3D-PDR code
(Dasyra et al. 2022), which self-consistently calculates the formation and destruction rate of molecules (Bisbas et al.
2015, 2017). For the CO, an increase of the cosmic ray ionization rate from about 10−16 s−1 to 10−14 s−1 was found
plausible to reproduce the millimeter data. In the optical, a significant shock contribution was found necessary by
Fonseca-Faria et al. (2023) to explain some high [Fe VII]/Hβ and [Fe X]/Hβ seen in MUSE data.
Both mechanisms are likely to be at play, but shocks seem to be of major importance according to the new H2
data for two reasons. The first reason has to do with the molecular hydrogen cooling rate. If we add the fluxes of the
H2 lines in the MIRI data that are primarily responsible for the radiative cooling, i.e., including S(1), S(2), and S(3)
(e.g., Ogle et al. 2010; Nesvadba et al. 2010), then we get a total luminosity of 9.2×1040 erg/s. By further adding the
flux of the S(0) line from the Spitzer IRS data, 4.6×10−17 W m−2 (Guillard et al. 2012), then the luminosity reaches
1.0×1041 erg/s. Dividing this luminosity with a total mass of 1.0×107 M⊙ (see subsection 4.3), we get a cooling rate
of 1.7×10−23 erg/s per molecule. Assuming, per Nesvadba et al. (2010), ∼12eV per ionization per molecule by cosmic
rays, this would require a cosmic ray ionization rate ζCR of order 10−12 s−1 to counterbalance the cooling. Even
though the actual ζCR could be up to an order of magnitude higher if a more massive reservoir could be detected from
similar S(0) observations, (see again subsection 4.3), this number would still remain high compared to the Galaxy, to
local AGN, and to the appropriate ζCR values found from the CO modeling (Dasyra et al. 2022).
Figure 14. Upper panels: H2 excitation temperature maps. All temperatures are in units of K. Crosses mark the centers of
the indicative regions (of 1′′ diameter) in which line fluxes were measured for the excitation diagram. The central three crosses
are identical to those of all previous Figures. The cross at the SE EELR corresponds to the location of the b1 bow shock. The
Tex is shown for the integrated line flux over all velocities, unless otherwise noted (i.e., except for the biconical outflow, shown
in the third column). Lower panels: H2 excitation diagram of the above-shown indicative regions.
Figure 15. H2 mass map, in units of solar masses, calculated for the coldest detected gas, probed by S(2) and S(1).
The second reason has to do with the detection of bow-shock-shaped features and outflowing gas well outside of the
8 GHz and 17.8 GHz emission contours, 1-1.5 kpc away from the nucleus (while the presently detectable radio lobes
reach only up to ∼500pc from the center). These features are compatible with a scenario in which the molecular gas
has been dragged out by previous episodes of the radio jet, which was more extended or bright back then compared to
the present-day radio jet. It is indeed frequent to observe recurrent radio activity in radio-loud objects (e.g. Owsianik
et al. 1998; Kuzmicz et al. 2017; Shabala et al. 2020), and statistics on a large number of radio sources reveal a short
duty cycle for radio jets, which could be explained by feedback-regulated accretion: in more than 10% of the cases,
there is evidence of radio lobe remnants, with a restarted activity in the center and with timescales between the two
jet activities of about 1-10 Myr. In this scenario, the H2 gas excitation has to have lasted longer, indicatively for
timescales of a few Myrs, after the jet passed from a region or faded. Indeed, the heating of the molecular gas could
be prolonged through shocks and the turbulence that they drive, with a dissipation timescale of the order of tens of
Myrs, as the turbulent energy cascade and dissipation happens for many orders of magnitude in spatial scale (e.g.
Ballesteros-Paredes et al. 1999; Guillard et al. 2009; Appleton et al. 2023; see also Le Bourlot et al. 1999 for the cooling
function of the H2 ).
In the center, i.e., in the region with detectable radio emission, the gas excitation can be elevated even in the case of
shocks alone, because of superposition. Several shocks may be traveling in the ISM at different directions and velocities
thanks to the jet scattering at various locations and thanks to the propagation of a slower moving, mass loaded gas
cocoon that is linked to the heavier outflow (Morganti et al. 2015). If shocks significantly contribute to the heating,
the gas should also be more turbulent, showing broader line profiles in regions where the gas is hotter. Indeed, this is
what we observe by comparison with the velocity dispersion maps of Fig. 8.
The most appropriate scenario or combination of scenarios per region will be addressed in a a forthcoming paper
using appropriate shock models that include cosmic ray heating, far UV photon heating, and (AGN or jet related)
X-ray heating. The latter heating mechanism will also be taken into account, even though it was previously deemed
to be rather weak at the position of the radio lobes: given the observed X-ray emission (Travascio et al. 2011), Dasyra
et al. (2022) calculated that a heating rate of 10−20 erg s−1 cm−3 , which would matter for the CO excitation, could
only be attained at the NW radio lobe if all clouds were near (e.g., within several parsecs of) a single X-ray source of
∼1039 erg s−1 . We further add here that a substantial increase in the stellar far UV radiation would not be able to
reproduce the CO emission either, so it is not a viable scenario by itself.
For the biconical outflow, seen in the west part of this region for 100<V<450 km s−1 , the same mechanisms may
also be contributing to the gas excitation. The biconical outflow’s excitation temperature also shows an increase in
the S(3) and S(5) based Tex map, in which this outflow is clearly detected. There, Tex of 800−1100 K are detected
instead of 650-700 K further out. These temperatures may be overestimated though, as the upper energy level (J=7)
of S(5) could be non-thermalized: the critical density of this transition is 3.5×105 cm−3 (Le Bourlot et al. 1999) and
fluorescence could then be contributing to the observed line emission.
Both in the vicinity of the radio lobes and in the biconical outflow there is a flux ratio of the total H2 emission (in
all lines) over the total PAH emission (in the two detected complexes) that is elevated by a factor of 5-10 compared to
regions further out, in the EELR. Even though this result could be reflecting a potential destruction of the PAHs by
the jets, and even though the PAH emission is rather weak in the spectrum of Fig. 2, such an interpretation is contra-
dicted by the finding that the PAH emission is bright along the jet axis (see also Diamond-Stanic & Rieke 2010; Ogle
et al. 2024). The PAH properties and excitation will too be discussed in a future paper, in the light of the gas excitation.
4.3. The molecular gas mass

We use the Tex map from S(1) and S(2) to determine the mass in each position of the galaxy, keeping in mind
that this Tex indeed describes a (near) LTE temperature up to J=5. Combining this temperature with the S(1)
luminosity map, we obtain the mass map shown in Fig. 15. The integrated mass in the FOV, M H2 is 1.0×107 M⊙ . An
ortho-to-para ratio of 3 is assumed for this calculation, which is reasonable for temperatures >200 K (e.g., Rigopoulou
et al. 2002). A more massive reservoir of H2 gas does plausibly exist (Guillard et al. 2012), but its proper mass
computation would require similar S(0) observations, as the mass measurement is highly sensitive to the Tex spatial
distribution. For comparison, in a similar aperture, the cold H2 mass from ALMA C0(1-0) data is 8.2(±0.6)×108 M⊙ ,
when plugging the CO(1-0) flux (30.6±2.4 Jy km s−1 ; Dasyra et al. 2022) and a Galactic CO luminosity to H2 mass
factor αCO =4.6 M⊙ K−1 km−1 s pc−2 (Bolatto et al. 2013) into Equation 3 of Solomon et al. (1997).
Concerning individual regions of interest, the bow-shock shaped structure contains 8.8×105 M⊙ of warm molecular
gas, while the mass around the NW radio lobe is similar, 7.8×105 M⊙ . Of the latter mass, 60(±14)% is linked to the
outflow in the H2 , as indicated by spectral fitting with a double Gaussian. The computation of the mass flow rate of
the accelerated gas requires knowledge of the deprojected velocity, the outflow geometry (Lutz et al. 2019; Veilleux et
al. 2020), and the exact outflow starting point(s) − thus the distance d traveled by the outflow(s). We performed an
order of magnitude calculation, assuming that the outflow(s) started locally, upon impact of the radio jet with clouds
near the lobe, for d in the range 60-250 pc (i.e., the difference in the spatial offset between the CO(1–0) peak and the
CO(4–3) peak in the ALMA data and the difference between the clump c2 and the NW lobe in the present JWST data),
and for an indicative velocity of 100 km s−1 . Then, the mass flow rate within 1′′ from the NW radio lobe, computed as
M H2 Vd−1 , is in the range 0.1−1 M⊙ yr−1 . Similarly, within 1′′ from the SE radio lobe, the mass is 7.4×105 M⊙ , the
outflow fraction is 40(±10)%, and the contribution to the total mass flow rate is just a little lower than that of the
NW lobe. Still, these numbers are low enough that the mass flow rate is dominated by the CO and HI emitting gas
(Morganti et al. 2007, 2015; Dasyra et al. 2016), even if some H2 were to be unaccounted for due to the lack of S(0) data.
5. CONCLUSIONS
We obtained JWST MIRI data of the nucleus of the nearby galaxy IC5063, in which a jet is known to interact with
dense gas clouds and drive multi-phase outflows, changing the gas excitation and cloud stability conditions. Our new
JWST data provided IFU datacubes of the inner ∼3×2kpc2 of the galaxy (in the common FOV of all observations),
and revealed the following findings.
• An unresolved source at the galaxy’s nucleus emits a bright MIR continuum, radially limited to 50pc, that can
be attributed to the torus with potential contribution from the radio core.
• Thirty nine spectral lines are detected on top of the continuum. Of those, the best tracers of the gaseous disk
are the low-J H2 lines, (0-0) S(1) and S(2). Instead, the ionized gas traces closely the jet trail and the optical
EELR. Higher ionization potential species tend to have a more nucleated emission. Given its IP and the spatial
resolution of MIRI at its wavelength, [Mg V] is the best tracer of the jet trail.
• Outflows are identified in more than ten discrete locations of IC5063. These include the vicinity of the two radio
lobes, the nucleus, and a biconical structure perpendicular to the jet trail, which host previously known outflows.
Additionally, new outflows are detected at points near or intermediate to the above, at the outer edges of the
EELR, in the interior of the EELR, and in an ionized gas bubble. Overall, outflows and turbulent kinematics are
seen in ∼2/3rds of the common FOV of our observations. The outflow velocities reach and occasionally exceed
1000 km s−1 for the ionized gas. In the bright [S IV] line, within ∼600 pc of the nucleus, the outflow velocities
often exceed the local escape velocity. For the molecular gas, the observed velocities are lower, typically not
exceeding the escape velocity.
• For both gas phases, greater outflow velocities are observed for higher ionization/excitation lines near the radio
lobes. This result indicates that the outflows are stratified, with the higher ionization/excitation gas being in
closer proximity to the power source (i.e., the jet).
• Newly detected bow shocks are seen in H2 lines in regions without significant radio emission, further away from
the nucleus than the radio lobes. Their presence suggests that a jet has propagated beyond what radio images
indicate. Located near the edge of the SE EELR, these bow shocks indicate that the jet got diverted or scattered
at the radio lobe near the SE EELR base, that it propagated along the EELR edge, and that it altered the gas
distribution, kinematics, and ionization there. Outflowing gas is detected all along this EELR edge, for ∼1 kpc,
including at the bow shock locations. Shock-driven turbulence is the most likely warm H2 excitation mechanism
in the bow shocks, as turbulent dissipation happens at much slower timescales than the passage of a relativistic
jet.
• The most highly excited warm H2 gas can be found where the most severe jet-cloud interactions take place: at
the radio lobes, in the biconical outflow, and potentially at the nucleus. The entire region with radio emission
shows a (step-function) increase of at least 100K in the excitation temperature compared to regions without
radio emission, such as the EELR interior. More jet-related gas excitation mechanisms must therefore operate
in that region, including cosmic rays and numerous shocks traveling at different directions/velocities.
• The mass of the warm H2 (>200K) is 1.0×107 M⊙ , as computed from the S(1) line emission map and the Tex
map based on S(1) and S(2) . This value is nearly two orders of magnitude below that of the cold H2 gas probed
by CO lines. The molecular gas masses of both the disk and the outflow are dominated by the cold, CO-emitting
gas component.
Acknowledgements: This work is based on observations made with the NASA/ESA/CSA JWST. The data were
obtained from the Mikulski Archive for Space Telescopes at the Space Telescope Science Institute.KMD would like to
thank her colleagues at the National and Kapodistrian University of Athens for the sabbatical that was crucial for
the preparation of this paper in a time-critical manner. KMD would also like to thank several members of the JWST
staff, namely D. Law for a lengthy and detailed communication on the flux calibration, as well as the ESA/STScI
team members M. Garcia-Marin, S. Kendrew, and K. Larson for pertinent discussions. US-based authors acknowledge
support by NASA under the JWST data analysis grant JWST-GO-02004.002-A. GFP acknowledges support by the
European Research Council advanced grant “M2FINDERS - Mapping Magnetic Fields with INterferometry Down
to Event hoRizon Scales” (Grant No. 101018682). JAFO acknowledges financial support by the Spanish Ministry of
Science and Innovation (MCIN/AEI/10.13039/501100011033), by “ERDF A way of making Europe” and by “European
Union NextGenerationEU/PRTR” through the grants PID2021-124918NB-C44 and CNS2023-145339; MCIN and the
European Union – NextGenerationEU through the Recovery and Resilience Facility project ICTS-MRR-2021-03-
CEFCA.
REFERENCES
Aalto, S., Costagliola, F., Muller, S., et al. 2016, A&A, 590, Donnan, C. T., McLeod, D. J., McLure, R. J., et al. 2023a,
73 MNRAS, 520, 4554
Aalto, S., Falstad, N., Muller, S. et al. 2016, A&A, 640, 104 Donnan, C. T., McLeod, D. J., Dunlop, J. S., et al. 2023b,
Appleton, P. N., Guillard, P. , Emonts, B., et al. 2023, ApJ, MNRAS, 518, 6011
951, 104 Endsley, R., Stark, D. P., Fan, X. et al. 2022, MNRAS, 512,
Argyriou, I., Glasse, A., Law, D.m et al. 2023, A&A, 675, 4248
111 Esparza-Arredondo, D., Gonzalez-Martin, O., Dultzin, D.,
Armus, L., Lai, T., U., V., Larson, K. L. et al. 2023, ApJ, et al. 2019, ApJ, 886, 125
942, L37 Ferland, G. J., Fabian, A. C., Hatch, N. A., et al. 2008,
Asmus, D, Hoenig, S. F. H., Gandhi, P., Smette, A., & MNRAS, 386, L72
Duschl, W. J. 2014, MNRAS, 439, 1648 Fernández-Ontiveros, J. A., Dasyra, K. M.,
Audibert, A., Dasyra, K. M., Papachristou, M. et al. 2022, Hatziminaoglou, E., et al. 2020 A&A, 633, 127
A&A, 668, 67 Finkelstein, S. L., Bagley, M. B., Haro, P. A., et al. 2022
Atek, H., Shuntov, M., Furtak, L. J., et al. 2023, MNRAS, Finkelstein, S. L., Bagley, M. B., Ferguson, H. C., et al.
519, 1201 2023, ApJL, 946, 13, ApJL, 940, L55
Ballesteros-Paredes, J., Vazquez-Semadeni, E., & Scalo, J. Fonseca-Faria, M. A., Rodriguez-Ardila, A., Contini, M.,
1999, ApJ, 515, 286 Dahmer-Hahn, L. G., & Morganti, R. 2023, MNRAS,
Barnouin, T., Marin, F., Lopez-Rodriguez, E., Huber, L., & 524, 143
Kishimoto, M. 2023, A&A, 678, 143 Juodzbalis I., Conselice, C. J., Singh, M. et al. (2023),
Barro, G., Perez-Gonzalez, Pablo G., Kocevski, D., et al. MNRAS, 525, 1353
(2023), ApJ, submitted Hernquist, L., 1990, ApJ, 356, 359
Bisbas, T. G., Bell, T. A., Viti, S., Yates, J. , Barlow, M. J. Gaibler, V., Khochfar, S., Krause, M., et al. 2012, MNRAS,
2012, MNRAS, 427, 2100 425, 438
Bisbas, T. G., Papadopoulos, P. P., & Viti, S. 2015, ApJ, Gaspar, A., Rieke, G. H., Guillard, P. et al. 2021, PASP,
803, 37 133, 4504
Bisbas, T. G., van Dishoeck, E. F., Papadopoulos, Padelis Guillard, P., Boulanger, F., Pineau Des Forets, G.,
P., et al. 2017, ApJ, 839, 90 Appleton, P. N. 2009, A&A, 502, 515
Bolatto, A. D., Wolfire, M. & Leroy, A. K. 2013, ARA&A, Guillard, P., Ogle, P., Emonts, B.H.C. et al. 2012, ApJ,
51, 207 747, 95
Bower, R. G., Benson, A. J., Malbon, R., et al. 2006, Harikane, Y., Ouchi, M., Oguri, M., et al. 2023, ApJS, 265,
MNRAS, 370, 645 5
Castellano, M., Fontana, A., Treu, T., et al. 2022, ApJL, Harikane, Y., Zhang, Y., Nakajima, K. et al. 2023, ApJ,
938 959, 39
Congiu, E., Contini, M., Ciroi, S., et al. 2017, MNRAS, Higdon, S. J. U., Armus, L., Higdon, J. L., Soifer, B. T., &
471, 562 Spoon, H. W. W. 2006, ApJ, 648, 323
Colina, L., Sparks, W. B., & Macchetto, F. 1991, ApJ, 370, Hönig, S.F., Kishimoto, M., Gandhi, P., et al. 2010, A&A
102 515, A23
Croton, D. J., Springel, V., White, S. D. M., et al. 2006, Holden, L., Tadhunter, C., Morganti, R., & Oosterloo, T.
MNRAS, 367, 864 2023, MNRAS 520, 1848
Dasyra, K. M., Ho, L., Netzer, H. et al. 2011, ApJ, 740, 94 Kocevski, D. D., Masafusa O., Kohei I., et al. 2023 ApJL,
Dasyra, K. M., & Combes, F. 2011, A&A, 533, L10 954, 4
Dasyra, K. M., Bostrom, A. C., Combes, F., & Vlahakis, N. Kulkarni, V. P., Calzetti, D., Bergeron, L., et al. 1998,
2015, ApJ, 815, 34 ApJL, 492, 121
Dasyra, K. M., Combes, F., Oosterloo. T. et al. 2016, A&A, Kuzmicz, A., Jamrozy, M., Koziel-Wierzbowska, D.,
595, L7 Weżgowiec, M. 2017, MNRAS,471, 3806
Dasyra, K. M., Paraschos, G. F., Bisbas, T. G., Combes, Labbe, I., Greene, J. E., Bezanson, R. et al. 2023, ApJ,
F., Fernández-Ontiveros, J. A. 2022, Nature Astronomy, submitted
6, 1077 Labrides, E., Chiaberge, M., Long, A. S. et al. 2024, ApJL,
Diamond-Stanic, A. M., & Rieke, G. H. 2010, ApJ, 724 961, 25
Larson, R., Finkelstein, S. L., Kocevski, D. D. et al. 2023, Owsianik, I., Conway, J. E., & Polatidis, A. G. 1998, A&A,
ApJL, 953, 29 336, L37
Le Bourlot, J., Pineau des Forêts, G., & Flower, D. R. Padovani, M., Bialy, S., Galli, D., et al. 2022, A&A, 658,
1999, MNRAS, 305, 802 189
Leftley, J. H., Nesvadba, N. P. H., Bicknell, G., et al. 2024, Panuzzo, P., Rampazzo, R., Bressan, A. et al. 2011, A&A
A&A, in press 528, 10
Leung, G., C. K., Bagley, M. B., Finkelstein, S. L. et al. Papachristou, M., Dasyra, K. M., Fernandez-Ontiveros, J.
2023, ApJL, ApJ, 954, 46 A., et al. 2023, A&A, 679, 115
Luridiana, V., Morisset, C., Shaw, R. 2015, A&A,573, 42 Pereira-Santaella, M., Alvarez-Marquez, J., Garcia-Bernete,
I. et al. 2022, A&A, 665, L11
Lutz, D., Sturm, E., Janssen, A., et al. 2019, A&A, 633, 134
Lyu, J., Alberts, S., Rieke, G. H., et al. 2023, ApJ, Rampazzo, R., Panuzzo, P., Vega, O. et al. 2013, MNRAS
submitted, arXiv:2310.12330 432, 374
Rigopoulou, D., Kunze, D., Lutz, D., 2002, A&A, 389, 374
Maiolino, R., Ubler, H, Perna, M. et al. 2024, A&A, in press
Roussel, H., Helou, G., Hollenbach, D. J., et al. 2007, ApJ,
Maiolino, R., Scholtz, J., Curtis-Lake, E., et al. 2023, A&A,
669, 959
submitted, arXiv:2308.01230
Rupke, D. S. N., Wylezalek, D., Zakamska, N. L., et al.
Maksym, W. P., Schmidt, J., Keel, W.C. et al. 2020, ApJL,
2023, ApJ, 953, L26
902, 18
Shabala, S. S., Jurlin, N., Morganti, R., et al. 2020,
Morganti, R., Oosterloo, T., & Tsvetanov, Z., 1998 AJ,
MNRAS, 496, 1706
115, 915
Solomon, P., Downes, D., Radford, S., et al. 1997, ApJ,
Morganti, R., Holt, J., Saripalli, L., et al. 2007, A&A, 476, 478, 144
735 Tadhunter, C., Morganti, R., Rose, M., et al. 2014, Nature,
Morganti, R., Oosterloo, O, Oonk, R. J. B., et al., 2015 511, 440
A&A, 580,1 Tommasin, S, Spinoglio, L. Malkan, M., & Fazio, G. 2010,
Mingozzi, M., Cresci, G., Venturi, G., et al. 2019, A&A, ApJ, 709, 1257
622, 146 Travascio, A., Fabbiano, G., Paggi, A., et al. 2021, ApJ,
Mukherjee, D., Wagner, A. Y., Bicknell, G. V., et al. 2018, 921, 129.
MNRAS, 476, 80 Vayner, A., Zakamska, N. L., Ishikawa, Y. et al. 2024, ApJ,
Naidu, R. P., Oesch, P. A., van Dokkum, P., et al. 2022a, 960, 126
ApJL, 940 Veilleux, S., Maiolino, R., Bolatto, A., Aalto, S., 2020,
Nesvadba, N. P. H., Boulanger, F., Salome, P., et al. 2010, A&ARv, 28, 2
A&A, 521, A65 Venturi, G., Cresci, G., Marconi, A., et al. 2021, A&A 648,
Nicastro F., Martocchia A., & Matt G., 2003, ApJ, 589, L13 A17
Ogle, P., Boulanger, F., Guillard, P. et al. 2010, ApJ, 724, Wagner, A. Y., & Bicknell, G. V. 2011, ApJ, 728, 29
1193 Wagner, A. Y., Umemura, M., & Bicknell, G. V. 2013, ApJ,
Ogle, P. M., Lopez, I. E., Reynaldi, V., et al. 2024, ApJ, 763, L18
962, 196 Wagner, S. J. & Appenzeller, I. 1989, A&A, 225, L13
Oosterloo, T. A.; Morganti, R.; Tzioumis, A., et al. 2000, Wells, M., Pel, J.-W., Glasse, A., et al. 2015, PASP, 127,
646
AJ, 119, 2085
Yan, H., Cohen, S., Windhorst, R. A., 2023 ApJL 942 L8
Oosterloo, T., Raymond Oonk, J. B., Morganti, R., et al.
Zubovas, K. & King, A. 2014, MNRAS, 439, 400
2017, A&A, 608, A38
APPENDIX
A. NOTES ON THE INSTRUMENT’S BEHAVIOUR IN THE PRESENCE OF BRIGHT LINE EMISSION

A.1. Data showing the cruciform artifact
Besides the PSF residuals that are related to a bright continuum source, there is an artifact that can affect any
part of the detectors that suffers from the leakage of photons from a bright emission line along the x axis of the
raw data. The issue is reported and depicted for the imager in the JWST user documentation webpage1 . It is
the so-called cruciform artifact. It originates from internal scattering of the light within the MIRI detectors, in
particular at lower wavelengths, which have a <27% quantum efficiency. Photons that are neither absorbed nor able
to escape (in particular for outer PSF regions) are diffracted by the lattice and cause straight-light patterns (Gaspar
et al. 2021; Argyriou et al. 2023). This was occasionally clearly seen in our data, as stripes oriented north-south
(e.g., in the reshifted [Ne III] and [Ne V] emission shown in Fig. 16). The cruciform was more easily seen for bright
emission lines. Bright lines contaminate neighboring detector slices at a constant y-axis pixel. But both because of
the detector distortion and also because neighboring detector slices do not correspond to neighboring sky positions,
the contamination appears at different wavelengths from the emission line peak wavelength and in various positions
in the reconstructed cubes’ FOV. Moreover, because of diffraction, the stray light can either be seen as a stripe, or as
an Airy diffraction pattern (see the [Ne VI] emission in Fig. 16). The pipeline, by default, automatically subtracts a
(straight line emission) model in channels 1 and 2, in an attempt to remove this stray light. However, because of the
detector distortion, signal may be oversubtracted from some regions and undersubtracted in other regions. This leads
to regions of images that appear to have cavities (e.g., [Ne III] in Fig. 16) and/or strong absorption (as in all panels
of Fig. 17), that can contain emission blobs within the absorption area. This can be confused for real absorption in
case there is a bright, warm background emission source such as a jet’s synchrotron in the background. Or it could
be confused for a P-Cygni or inverse P-cygni profile. In our case, there are winds (Dasyra et al. 2015), but there is no
detected synchrotron emission in the regions where the absorption is seen. (Note that the images and spectra shown
in Fig. 17 are continuum subtracted.) Therefore, the absorption would have to be tracing outflows of cold, yet highly
ionized gas that is absorbing the emission of hot dust in the background - a rather unusual configuration. However,
we report the presence of the artifact for demonstration purposes.
Figure 16. Emission due to the cruciform artifact.
A.2. Data with signal potentially linked to the cruciform artifact

Because of this artifact and PSF residuals, we occasionally see signal at low or high velocities that cannot be certainly
determined to be true. Examples of regions (except the nucleus, radio lobes, and already reported knots) where more
outflows could potentially exist are shown in Fig. 18. The presence of outflows in these regions needs to be tested
1 https://jwst-docs.stsci.edu/known-issues-with-jwst-data/miri-known-issues#MIRIKnownIssues-cruciformCruciformartifact
with other data or MIRI data of different setup. A most striking feature is a filament that starts from the NW radio
lobe, then tilts and continues in a north-south orientation. It is presented in the selected [Ar II] and H2 S(2) images
of Fig. 18, but it is present in a plethora of lines.
Figure 17. Absorption due to the cruciform artifact.
Figure 18. Emission that could be associated with regions of interest, but that cannot be confirmed to be real due to the
cruciform artifact and PSF residuals.
A.3. Nuclear PSF

The image of the continuum for each of the four MIRI channels is shown in Fig. 19. The figure reveals the considerable
domination of the nuclear PSF over any other continuum source in the FOV.
Figure 19. Continuum images in the medium band of each channel (i.e., ch1 to ch4 from left to right).

2406 03218v1

Uploaded by

Copyright:

Available Formats

2406 03218v1

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2406 03218v1

Uploaded by

Copyright:

Available Formats

Draft version June 6, 2024

Typeset using LATEX default style in AASTeX631

A case study of gas impacted by black-hole jets with the JWST:

3.2. Regions influenced by the jet and outflows

3.3. Density of the ionized gas

102 104 106 108

4. DISCUSSION: THE EXCITATION MECHANISM(S) OF THE GAS

Table 1. Line properties

Line name λ f(a ) σ

4.2. The molecular gas excitation

4.3. The molecular gas mass

A. NOTES ON THE INSTRUMENT’S BEHAVIOUR IN THE PRESENCE OF BRIGHT LINE EMISSION

Figure 16. Emission due to the cruciform artifact.

A.2. Data with signal potentially linked to the cruciform artifact

Figure 17. Absorption due to the cruciform artifact.

A.3. Nuclear PSF

You might also like