Ram
pressure diagnostics using polarized emission
Team:
Bernd Vollmer, Observatoire de Strasbourg, France
(PI)
Marian Soida, Jagiellonian University, Krakov, Poland
Marek Urbanik, Jagiellonian University, Krakov, Poland
Rainer Beck, Max-Planck-Institut für Radioastronomie, Bonn, Germany
Krzysztof T. Chyzy, Jagiellonian University, Krakov, Poland
Katarzyna Otmianowska-Mazur, Jagiellonian University, Krakov, Poland
Jeffrey Kenney, Yale University, New Haven, USA
Jacqueline van Gorkom, Columbia University, New York, USA
Aeree Chung, National Radio Astronomy Observatory, USA
Marek Wezgowiec, Ruhr Universitat Bochum, Germany
Ram pressure stripping in the Virgo cluster
The mapping of the gas content of spiral galaxies in the Virgo cluster
showed that the HI disks of cluster spirals are disturbed and
considerably reduced (Cayatte et al. 1990, 1994). These
observational results indicate that the gas removal due to the rapid
motion of the galaxy within the intracluster medium (ICM) (ram pressure
stripping; Gunn & Gott 1972) is
responsible for the HI deficiency and the disturbed gas disks of the
cluster spirals. Nevertheless, it is still an open question where and
when these galaxies lose their gas.
- Do spiral
galaxies lose their gas during impulsive events in the Virgo cluster
center (D <
0.5Mpc) as proposed by Vollmer et al. (2001) or can ram pressure be
more efficientthan
we thought at intermediate distances (~1Mpc) from the cluster center as
we observed it
for 7 extended (or long) HI tailed galaxies within the VIVA survey
(Chung et al. 2007) or
for NGC 4522?
- NGC 4522 is a
special case, because it suggests a rapidly moving intracluster medium due to the infall of the M49 group of
galaxies into the Virgo cluster (Kenney et al. 2004, Vollmer et al.
2006). Is NGC 4522 a unique case?
- Does gas infall
after peak ram pressure really occur, as suggested by the Vollmer et
al. (2001) models?
- Is the HI
deficiency of galaxies far away from the Virgo cluster center (5-10~Mpc, Solanes et al.
2002) due to ram pressure stripping?
During a ram pressure
stripping event the ISM of the galaxy is compressed by the ram pressure
of the intracluster medium. After the galaxy's passage through the
cluster core ram pressure decreases considerably and the ISM that has
not been accelerated to the escape velocity possibly falls back onto
the galactic disk. The backfalling gas causes shear motions. Both,
compression and shear motions, cannot be easily detected in an HI
velocity field. However, polarized radio continuum emission is very
sensitive to these motions, because they generate anisotropic
fields from isotropic random fields and they enhance locally the
ordered magnetic field on large scales. Thus, whenever there is
compression or shear the polarized radio continuum emission
lights up. This emission must be observed at a wavelength small enough
to avoid significant Faraday
depolarization.
The combination of the HI gas
distribution, velocity field and
the distribution of the polarized radio emission enables us to
diagnostic if and when ram
pressure affected the observed galaxy (see Kenney et al. 2004,
Vollmer et al. 2004b, and Vollmer et al. 2006 for NGC 4522, Soida et
al. 2006 for NGC 4654, and Vollmer et al. 2008 for NGC 4501).
Why polarized radio continuum emission
Polarized radio continuum emission (PI) is sensitive to the
density of cosmic ray electrons and,
most importantly, on the ordered magnetic field on large scales. In an
exploring work (Otmianowska-Mazur & Vollmer 2003) we solved the
induction equation on the velocity fields given by the above described
numerical simulations to calculate the evolution of the ordered
magnetic field. With an assumed distribution of cosmic ray electrons
the evolution of the PI was determined. We showed that the PI increases
whenever there is compression or shear. Both phenomenon occur during a
ram pressure stripping event. During ram pressure maximum the gas is
compressed; after the ram pressure maximum (when the galaxy is leaving
the cluster core) the resettlement of the gas causes shear. The best
example for the synergy between the VLA PI observations and our MHD
modelling is NGC 4522. Kenney et al. (2004) detected high column
density, extraplanar HI to the west of the almost edge-on galactic
disk. With deep 6cm observations of the polarized emission (Vollmer et
al. 2004) we directly detected the compression region on the opposite
side of the galaxy. A dynamical and MHD modelling show that this galaxy
is close to peak ram pressure despite its large distance from the
cluster center (Vollmer et al. 2006). This implies that the
intracluster gas is rapidly moving towards the location of NGC 4522.
This intracluster gas is most probably associated with the infalling
M49 group of galaxies.

NGC 4522. Left panel: greyscale: optical R band image.
Contours:
HI gas distribution (from Kenney et al. 2004).
Right panel: greyscale: HI gas distribution.
Contours: polarized radio emission at 6cm (from Vollmer et al. 2004).
This kind of dynamical and MHD modelling was also applied to NGC
4654 (Soida et al. 2006).
The dynamical modelling of the HI data (Vollmer 2003) has shown that
this galaxy underwent a tidal encounter ~500Myr ago and is now affected
by a weak ram pressure stripping. The MHD modelling of the tidal
interaction alone (GR model), ram pressure stripping alone (RPS model),
and a mixed interaction (GRPS; Soida et al. 2006) confirmed the earlier
result.

NGC 4654: Greyscale: HI distribution. Contours: polarized
radio
continuum emission. GR model: gravitational interaction alone.
RPS model: ram pressure stripping alone. GRPS model: mixed
interaction. Right panel: VLA HI and 6cm continuum observations
(Soida et al. 2006).
For NGC 4501 only the 6cm polarized radio continuum observations
and the MHD modelling show unambiguously that this galaxy undergoes
active ram pressure
stripping. The ridge of polarized 6cm radio continuum emission at the
outer southwestern disk is a clear sign of compression. This
compression was not detectable in the velocity field. The HI
overdensity and sharp edge suggested a compression, but only the
polarized 6~cm radio continuum emission
could confirm it.

NGC 4501: Greyscale: HI distribution. Contours:
polarized
radio continuum emission. Left panel: VLA observations.
Middle and right panel: ram pressure models with different
galaxy orbits (Vollmer et al. 2008).
Previous VLA polarized radio continuum observations
In a pilot study we observed 8 Virgo spiral galaxies at 20cm and
6cm in polarized emission.
All observed Virgo spiral galaxies (except NGC~4321) show a strongly
asymmetric distribution of polarized radio continuum emission with
elongated ridges located in the outer galactic disk. Such polarized
emission ridges are also observed in two other Virgo
spiral galaxies: NGC 4254 (Chyzy et al. 2007) and NGC 4522 (Vollmer et
al. 2004). Based on this sample of 10 Virgo spiral galaxies observed
with the VLA,
we conclude that the 6cm polarized radio continuum emission
distribution in Virgo spiral galaxies is different from that of field
spiral galaxies (Beck 2005), in the sense that the
distribution is strongly asymmetric with elongated ridges in the outer
part of the galactic disks. Since all asymmetric emission ridges are
located in the outer galactic
disks, they are most likely due to external influences of the cluster
environment on the galaxies. These influences can be of tidal or
hydrodynamic nature. Thus the distribution of polarized radio continuum
emission contains important information about the velocity distortions
caused by the interaction of a spiral galaxy with its cluster
environment. In addition, this information is complementary to that of
CO or HI data cubes, because we can observe shear or compression caused
by the
velocity components in the plane of the sky. It is very difficult, if
not impossible, to predict the distribution of polarized radio
continuum emission on the basis of CO or HI observations, because of
the complex evolution of the magnetic field (induction equation) and
beam depolarization effects. It is therefore necessary to do detailed
MHD modelling of individual galaxies for direct comparison with
observations.

6cm polarized intensities as contours and the vectors of
the
magnetic field uncorrected for Faraday rotation on VLA HI
images. The size of the lines is proportional to the intensity
of the polarized emission (Vollmer et al. 2008).
Target selection
The target galaxies can be divided into 4
categories according to the
following criteria:
(i) galaxy pairs,
(ii) galaxies which are likely
affected by ram pressure (distance to M 87 D < 3deg),
(iii) galaxies at the periphery (4deg
< D < 8deg) of the cluster,
(iv) galaxies far away from the Virgo
cluster (D > 8deg).

Virgo cluster X-ray emission (ROSAT) of the intracluster
medium and HI
emission of the proposed sample of Virgo spiral galaxies
(from VIVA; image credit: A. Chung, NRAO).