Thesis

OBSERVATOIRE DE PARIS
ECOLE DOCTORALE
ASTRONOMIE ET ASTROPHYSIQUE D’ILE-DE-FRANCE
Doctorat
ASTRONOMIE ET ASTROPHYSIQUE
AUTHOR: CELIA ANAHI VERDUGO SALGADO
STAR FORMATION IN LOW GAS DENSITY AND LOW

METALLICITY ENVIRONMENTS
Thesis supervised by FRANÇOISE COMBES
Defended on July 21st 2015, in Paris
Jury:
Eva Schinnerer (examiner)
Samuel Boissier (examiner)
Pierre-Alain Duc
Nicolas Bouché
Matthew Lehnert
David Valls-Gabaud
OBSERVATOIRE DE PARIS
ECOLE DOCTORALE
ASTRONOMIE ET ASTROPHYSIQUE D’ILE-DE-FRANCE
Doctorat
ASTRONOMIE ET ASTROPHYSIQUE
AUTEUR: CELIA ANAHI VERDUGO SALGADO
FORMATION DES ÉTOILES DANS LES

ENVIRONNEMENTS DE BASSE DENSITÉ DE GAZ ET
BASSE MÉTALLICITÉ
Thèse dirigée par FRANÇOISE COMBES
Soutenue à Paris le 21 Juillet 2015
Jury:
Eva Schinnerer (rapporteur)
Samuel Boissier (rapporteur)
Pierre-Alain Duc
Nicolas Bouché
Matthew Lehnert
David Valls-Gabaud
Abstract
In nearby galaxies, an empirical relation has been established between star forma-
tion and gas surface densities, the Kennicutt-Schmidt relation (K-S). The relation is
nearly linear when molecular gas (H2 ) is considered, while is less tight with atomic
hydrogen (HI). These low gas density regions are of a key importance in the field
of star formation, since the are also low metallicity environments, resembling the
conditions of a younger universe. This thesis summarizes the observational work
done with the IRAM 30mt telescope in two kinds of such regions: disk galaxies with
extended ultra-violet emission (XUV), and the interstellar medium going into the
hot intra-cluster medium (ICM) under ram-pressure stripping in the Virgo Cluster.
The GALEX telescope has unveiled in Far Ultra-Violet (FUV) star formation in the
outer parts of some disk galaxies that was not traced by Ha. To determine the pres-
ence of H2 and analyze the K-S relation in these regions, CO observations were done
in the outskirts of several XUV disk galaxies, finding both detections and upper
limits. These K-S relations showed a broken power law at low gas densities, below
the HI-H2 threshold. In the Virgo Cluster, similar CO observations were done along
the HI tidal arm connecting NGC4388 and M86, where no H2 is expected. Two
detections were found, showing very low star formation efficiencies (depleting less
than 0.1% of the gas reservoir per 108 yr), and showing again a discontinuity of the
K-S relation at low gas densities, probing that the process of gas consumption into
stars well known at high densities cannot be extrapolated to lower densities, and
that H2 can survive a certain time in the hostile ICM.
4
Abstract
Dans les galaxies proches la relation empirique de Kennicutt-Schmidt (K-S) a
été établie entre la densité surfacique de formation d’étoiles et la densité surfacique
de gaz. Elle est forte et presque linéaire quand on considère le gaz moléculaire
(H2 ) mais devient moins valable lorsque l’hydrogène n’est qu’atomique (HI). Ces
régions de densité de gaz faible sont d’une importance cruciale dans le domaine
de la formation d’étoiles, car elles possèdent les mêmes conditions que dans les
galaxies de Univers jeune, de faible métallicité. Cette thèse compile les données
observationnelles obtenues au télescope de 30m de l’IRAM sur deux types de régions
distinctes: les galaxies à disque ayant une émission ultra-violet étendue (XUV), et
le milieu interstellaire déplacé dans le milieu intra-amas (ICM) chaud sous l’effet
du balayage de la pression dynamique dans l’amas de la Vierge. Pour déterminer
la présence de molécules H2 et afin d’étudier la relation K-S dans ces régions, des
observations de CO ont été faites dans les parties externes des galaxies XUV, et
nous avons à la fois des détections et des limites supérieures. Les relations K-S ont
montré une loi de puissance discontinue pour des densités de gaz faibles, en dessous
du seuil de la transition de phase HI-H2 . Dans l’amas de la Vierge, des observations
similaires de CO ont été faites le long des bras de marée qui relient NGC4388 et
M86, où aucune présence de H2 n’est attendue. Nous avons effectué des détections,
montrant une efficacité de formation d’toiles très faible, et montrant de nouveau
une discontinuité de la relation K-S pour des densités de gaz faibles, prouvant que
les processus bien connus de consommation du gaz à forte densité ne peuvent être
généralisés aux densités plus faibles. D’autre part, nos résultats montrent que le gaz
H2 peut survivre un temps plus long que prévu dans l’ICM hostile.
5
Acknowledgements
Throughout these three years of PhD I have a lot of people to thank for, that
have made my life just wonderful. It wasn’t easy to spend such a long time abroad,
leaving friends, family... well, an entire life, on the other side of the ocean. At
moments I felt very lonely and homesick, but thanks to the unconditional devotion
of my loved ones in Chile, regardless of the distance, and also to the wonderful
people I met here, my situation transformed up-side-down and became one of the
most wonderful experiences in my life. I know I will never stop thanking you all for
everything, but this is a start.
First of all I would like to thank my family for always supporting me in every
decision I have made, especially during my career. Even when the path I decided to
take was so uncommon, I never felt anything else but support and pride from my
entire family, and I will always cherish that. I just hope someday I will be able to
do everything you have done for me in return.
And of course, many thanks are ought to Françoise, for the constant guidance
throughout this work, and the permanent availability for any question, any day at
any time. I feel extremely lucky to have had such a remarkable investigator as my
tutor, and I conclude my PhD with a great feeling of knowledge and achievement. I
especially thank her her kind comprehension and patience on every difficult aspect I
found in this period. And also part of my guidance, I’d like to thank Mónica Rubio
and Alberto Bolatto for all their kind help and support in my work throughout these
last five years.
I also want to thank my friends in Chile, my friends from childhood, that re-
gardless of the time or distance, never forget about me and always keep me in their
minds. Thank you so much for so many years of friendship, laughter and good times.
We have shared so much that we are soul mates for life.
But of course my experience here in France wouldn’t have been the same without
the friends I made here. A beautiful group of people, most of the from Chile, but
from other countries as well, that have become my family here. They are the main
reason of my happiness here, and I will never forget them. Some of them I’ll meet
again in Chile, and others will be forever in my heart.
And finally, but most important, I thank my boyfriend, my best friend and my
soul mate for making my life so wonderful. You have not only been my major pillar
during this thesis, but changed my life forever. You were my best friend and then
became my soul mate, always there for me, no matter what. Your faith in my doing
this successfully is amazing. No one understands like you the heavy road this has
been, and I’m sure we only have happiness ahead.
Additionally, special acknowledgments are ought to the IRAM staff for their
support during the observations for this thesis, and to both the Centre National
de la Recherche Scientifique (CNRS) and the Comisión Nacional de Investigación
Cientı́fica y Tecnológica (CONICYT) for supporting my financially during the three
years of my PhD, through agreement signed on December 11, 2007.
6
Contents
1 Introduction 12
1.1 Observing molecular gas . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.2 Molecular gas mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.3 Kennicutt-Schmidt relation . . . . . . . . . . . . . . . . . . . . . . . 17
1.4 Low gas density environments . . . . . . . . . . . . . . . . . . . . . . 20
1.5 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2 XUV 23
2.1 M63 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.1 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.1.4 Summary and conclusions . . . . . . . . . . . . . . . . . . . . 48
2.1.5 Publication . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2 2013 observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2.2.1 Data reduction and Results . . . . . . . . . . . . . . . . . . . 63
2.2.2 Stacking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.3 M83 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.3.1 Instrumentation and Observations . . . . . . . . . . . . . . . . 69
2.3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.3.3 Star formation rate . . . . . . . . . . . . . . . . . . . . . . . . 72
3 Ram-Pressure stripping 76
3.1 Virgo cluster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.2 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.3 Data Reduction and spectral line fitting . . . . . . . . . . . . . . . . 80
3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5 Star formation efficiency . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5.1 HI photometry . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.5.2 Hα data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.5.3 K-S relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.6 Publication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4 Summary and Conclusions 102

A.1 CO in low metallicity galaxies . . . . . . . . . . . . . . . . . . . . . . 108
A.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
7
A.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
A.1.3 Publication . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
8
List of Figures
1.1 H2 energy levels diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1.2 Kennicutt-Schmidt relation in nearby galaxies. . . . . . . . . . . . . . 18
1.3 IRAM 30-mt single dish radio telescope. . . . . . . . . . . . . . . . . 21
1.4 Atmospheric transmission at Pico Veleta. . . . . . . . . . . . . . . . . 22
2.1 NUV-FUV color image of M63. . . . . . . . . . . . . . . . . . . . . . 26

2.2 Observing pointings for M63. . . . . . . . . . . . . . . . . . . . . . . 28
2.3 Mosaic of CO spectra obtained along the major axis of M63. . . . . . 32
2.4 CO spectrum for UV region in M63. . . . . . . . . . . . . . . . . . . 33
2.5 Maps of complementary data for M63. . . . . . . . . . . . . . . . . . 35
2.6 Multiwavelength views of the UV region in M63. . . . . . . . . . . . . 36
2.7 Image processing in complementary data for M63. . . . . . . . . . . . 37
2.8 Radial profiles of CO observations and complementary data for M63. 40
2.9 Radial profiles of SFE and H2 /HI for M63. . . . . . . . . . . . . . . . 43
2.10 Comparison of radial profiles of ΣSFR , ΣHI and ΣH2 between this work
and Leroy et al. (2008). . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.11 HI map of M63. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.12 Kennicutt-Schmidt relations of M63. . . . . . . . . . . . . . . . . . . 46
2.13 XUV galaxies observed in 2013. . . . . . . . . . . . . . . . . . . . . . 64
2.14 Spectra results for XUV observatons 2013. . . . . . . . . . . . . . . . 66
2.15 CO stacked spectra for NGC3344. . . . . . . . . . . . . . . . . . . . . 68
2.16 ALMA observatory at Llano Chajnantor in Chile. . . . . . . . . . . . 69
2.17 Atmospheric transmissions and spectral bands for ALMA. . . . . . . 70
2.18 M83 UV map. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.19 ALMA CO(2–1) intensity for M83. . . . . . . . . . . . . . . . . . . . 72
2.20 Hint of a possible clump in the CO(2–1) data cube of M83. . . . . . . 73
2.21 ALMA continuum map for M83. . . . . . . . . . . . . . . . . . . . . . 74
2.22 Complementary data for M83. . . . . . . . . . . . . . . . . . . . . . . 74
2.23 ΣSFR maps for M83. . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.1 Hα + NII image of the Virgo cluster from Kenney et al. (2008). . . . 79
3.2 Targets observed in the Virgo Cluster. . . . . . . . . . . . . . . . . . 81
3.3 CO detection for NGC4388, Source-1and HaR-2. . . . . . . . . . . . . 83
3.4 CO detection for second run of observations in the Virgo Cluster . . . 84
3.5 HI column density map from Oosterloo & van Gorkom (2005) . . . . . 88
3.6 Hα map of the Virgo cluster from Kenney et al. (2008). . . . . . . . . 90
9
3.7 K-S relation for sources observed in the Virgo cluster. . . . . . . . . . 91
1 Color composite image of the WLM galaxy. . . . . . . . . . . . . . . 109

2 Example of a modified blackbody fit. . . . . . . . . . . . . . . . . . . 111
10
List of Tables
2.1 CO mapping along the major axis of the M63 disk. . . . . . . . . . . 30

2.2 CO mapping over the bright UV region at 1.36 r25 . . . . . . . . . . . 31
2.3 Public ancillary data for M63. . . . . . . . . . . . . . . . . . . . . . . 34
2.4 Photometry results obtained for the complementary data of M63 in
its radial cut. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.5 Photometry results obtained for the complementary data of M63 in
its UV region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6 Best-fitting bisector linear Kennicutt-Schmidt relations obtained for
M63 in log space. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.7 Galaxies selected for XUV observations in 2013. . . . . . . . . . . . . 64
2.8 Targets description for XUV observations 2013. . . . . . . . . . . . . 65
2.9 Line parameters for detections in XUV observations 2013. . . . . . . 65
2.10 3σ upper limits of CO(1–0) emission for sources with no detection in
2013 observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.1 Targets and Observations. . . . . . . . . . . . . . . . . . . . . . . . . 82

3.2 CO(1–0) upper limits at 3σ for sources with no detection. . . . . . . . 85
3.3 CO detections in the Virgo Cluster. . . . . . . . . . . . . . . . . . . . 86
3.4 K-S relation values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
11
Chapter 1
Introduction
Contents
1.1 Observing molecular gas . . . . . . . . . . . . . . . . . . . 15
1.2 Molecular gas mass . . . . . . . . . . . . . . . . . . . . . . 16
1.3 Kennicutt-Schmidt relation . . . . . . . . . . . . . . . . . 17
1.4 Low gas density environments . . . . . . . . . . . . . . . . 20
1.5 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . 21
The process in which stars are born in the Universe holds one of the key answers
to our origins. It sheds light on the different conditions needed for the formation
of all types of stars, including ours and our planetary system. However, it is also
the root of astrophysics on its largest scales, since it is a principal agent of galaxy
formation and evolution. This is why in astronomy, the field of star formation has
never lost interest throughout the decades.
Thanks to many years of research (Kennicutt & Evans 2012 and references
therein), we now know that, on large scales, star formation events are determined
by a hierarchy of physical processes (e,g. Elmegreen & Elmegreen 2014, Drazinos
et al. 2013) spanning a vast range of physical scales. We could roughly describe it as
a first accretion of gas onto disks from satellite object and the intergalactic medium
(megaparsec scale) (e.g Combes 2014, Bouché et al. 2013, Sánchez Almeida et al.
2014), followed by the cooling of this gas to form a cool neutral phase (kiloparsec
scale), then the formation of molecular clouds (tens of parsecs) (e.g Dobbs et al.
2014), the fragmentation and accretion of this molecular gas to form progressively
denser structures such as clumps (∼1 pc) (Zhang 2014) and cores (∼0.1 pc) (Gong
& Ostriker 2015, Zhang et al. 2015), to finally initiate the contraction of the cores
to form stars (solar radius) and planets (∼astronomical units). Because these pro-
cesses involve physical scales and local conditions so different from each other, they
have been studied by different teams who have specialized in each one of them, to
parameterize the best possible way these various processes.
Consequently, these groups have grown separately, and enormous efforts have
been made to bring closer the community studying individual star-forming regions
12
and stars in the Milky Way, and the large extragalactic community that attempts
to characterize the star formation process on galactic and cosmological scales.
Of course, the historical distance between these two fields lies on observational
restrictions. The key physical processes that determine how molecular clouds con-
tract and fragment into clumps and cores and finally clusters and individual stars
can be probed up close only in our Galaxy, and much of the progress in this sub-
ject has come from in-depth case studies of individual star forming regions. Such
detailed observations have been impossible to obtain outside our galaxy, even rela-
tively nearby galaxies (with the exception of the Magellanic Clouds). The spatial
resolution required for that is too high, and hence those observations have been
restricted to a large scale only.
However, the past decade has witnessed an unprecedented stream of new obser-
vational information on star formation on all scales, thanks to the very successful
space mission telescopes, like the Galaxy Evolution Explorer (GALEX), the Spitzer
Space Telescope, and the Herschel space Observatory, as well as the state-of-the-art
ground based telescopes, like the Atacama Large Millimeter Array (ALMA), the
Institut de Radioastronomie millimétrique (IRAM), or the Very Large Telescope
(VLT), just to name a few. These facilities have not only provided key and detailed
information on the physical processes leading to the formation of individual stars in
interstellar clouds, but also mapped for the first time entire galaxies and ensembles
of galaxies, denoting evidences of star formation processes going on, and even more
fascinating, the interactions of these processes with their extragalactic medium, and
how they influence the evolution of their host galaxies.
With this instrumentation, the extragalactic branch of the field has focused on
the collective effects of star formation, integrated over entire star-forming regions or,
often, over entire galaxies (e.g Gil de Paz et al. 2007a, Kennicutt et al. 2011, Dale
et al. 2009, Walter et al. 2008). This collective conversion of baryons from interstellar
gas to stars and the emergent radiation and mechanical energy from the stellar pop-
ulations is most relevant to the formation and evolution of galaxies. Consequently,
much of the empirical knowledge of star formation on these scales consists of scal-
ing laws and other parametric descriptions, in place of rigorous, physically based
characterizations. Improving our knowledge of large-scale star formation and its
attendant feedback processes is essential to understand the formation and evolution
of galaxies.
A lot of the work done in star formation beyond our Milky Way (MW), in the
nearby Universe, involves the analysis of different star formation tracers present in
local and normal spiral disk galaxies. The new generation of telescopes has provided
maps of entire galaxies at different wavelengths, including ultra-violet (UV), Hα,
atomic gas (HI), molecular gas (H2 ) and infrared. These emissions can be directly
correlated with the amount of ongoing star formation, and more precisely, quantify
it. It has been shown that in these galaxies the amount of star formation depends
directly on the amount of molecular gas present, in a more or less linear relation,
and the gas consumption times are fairly constant, between ∼2 and 3 Gyrs (Leroy
et al. 2008). But very little is known about the efficiency of this process on other
13
types of galaxies and environments, including irregular galaxies and mediums of
low gas density and/or low metallicity. Such regions seem more difficult to observe
since their emission is much colder and fainter than in normal galaxies, but they
are not devoided of star formation, and to understand in detail the processes they
go through and to quantify their efficiencies to form stars is of a key importance to
someday fully understand star formation at all its levels.
This thesis explores the extragalactic branch of star formation, focusing on the
regions of galaxies in the local universe, where star forming events can interact with
the surrounding environments of the host galaxies. Such regions involve particularly
the outer parts of nearby disk galaxies with evidences of recent star formation, and
the intergalactic medium of galaxies that belong to a cluster and interact with
each other. It is an observational work done with ground-based telescopes, and
complemented with space telescope data. The aim of this work is to shed light on
the star formation processes taking place in regions where it is more difficult to
form stars, due to extreme conditions, such as low metallicity, low gas density, and
physical interaction with the surrounding environments. The study of star formation
in these regions not only help us to constrain boundaries on the conditions needed
to form stars, but also help us to better understand the past of our cosmos, since
they resemble the conditions of a younger universe.
The thesis is structured as follows. The present chapter gives a description of
the typical regions observed in this work, characterized by their low gas densities
and low metallicities. It also explains how we observe these regions and how we can
estimate the amount of molecular gas out of which stars can form. The Kennicutt-
Schmidt relation is also introduced, to quantify the star formation rate (SFR) and
star formation efficiency (SFE) as a function of the gas density. Finally, the instru-
mentation used in this work is explained. Chapter 2 explains the first observational
work done for this thesis, on nearby galaxies with extended disks of ultra-violet
emission (XUV), including radio observations, and analysis with complementary
data. It contains the work done on the M63 galaxy, with the results of success-
ful carbon monoxide (CO) observations and the first Kennicutt-Schmidt relation
ever done on a XUV disk galaxy, with its corresponding publication in the A&A
Journal. It also encompasses more CO observations on other XUV galaxies, with
both the IRAM and ALMA telescopes, and their preliminary results. Chapter 3
describes the second part of this thesis, on a tail of gas stripped from the NGC4388
galaxy, going under ram-pressure stripping in the Virgo Cluster. It includes success-
ful CO detections in the intergalactic medium, were molecular gas is unexpected. A
Kennicutt-Schmidt analysis was also done in this part, with extremely low values
of star formation efficiencies, not reported before, and a submitted publication to
A&A. Chapter 4 presents the final analysis and conclusion of this thesis. Also, in the
appendix I present an additional work, that started during my master thesis, but
finished during the first year of this PhD thesis, on CO observations at extremely
low metallicities. It is the first CO detection ever found on a galaxy of such a low
metallicity (O/H abundance less than 20%). The successful work published in the
Nature Journal is included here.
14
1.1 Observing molecular gas
The best way to analyze the star formation activity in a region and quantify
its efficiency, is to measure the amount of gas available to be converted into stars,
which as we know, are formed inside molecular clouds. These regions are constituted
mainly of hydrogen (H), which at the typical low temperatures of molecular clouds
(∼10 K) tends to be molecular (i.e H2 ). Hence the molecular gas holds the key
answer to the efficiency of any star formation process, however, is extremely difficult
to observe.
Even though hydrogen is the most abundant element in the universe, observing
molecular hydrogen (H2 ) is highly difficult in the interstellar medium (ISM). For
H2 , the first excited state, the J = 1 rotational state, is 100 - 200 K above ground
state (Figure 1.1), whereas typical molecular clouds have temperatures of a few tens,
making this transition highly unlikely. Even more, since the H2 is a homonuclear
molecule, it is practically impossible to detect because it has no electric dipole
transitions that can change J by 1 (therefore no J = 1→0 emission). Instead, the
lowest-lying transition is the J = 2→0 quadrupole, very weak and unlikely, since the
J = 2 state is 511 K above ground state.
Figure 1.1: H2 energy levels diagram
The basic reason of why it is so difficult to observe H2 is its low mass (3.34×10−27
15
kg). For a quantum oscillator or rotor the level spacing varies with reduced mass as
m−1/2 , making it so difficult to go from one transition to another.
For these reasons we turn to other tracers to determine the amount of H2 . The
most common one is the CO molecule, thanks to its dipole moment and its bright
emission line. Carbon and oxygen are two of the most common elements in the ISM
after H and He, and since the CO molecule is much more massive than H2 , its lowest
rotational state is only 5.5 K above ground state, making CO molecules abundant
with bright emission lines, easy to detect. As a consequence, CO has long been used
as the preferred molecular gas tracer.
1.2 Molecular gas mass

As we know, stars form out of molecular gas, due to gravitational collapse inside
molecular clouds. Therefore, in star formation analysis, including star formation
rates (SFRs) and star formation efficiencies (SFEs), it is of crucial importance to
measure the amount of molecular gas available to form stars.
As explained in the previous section, it is extremely difficult to detect molecular
gas, i.e H2 , in low gas density environments and the cold ISM. In these cases we
need to turn to CO(1–0) emission to detect H2 and measure its mass.
The H2 column density, which corresponds to the amount of H2 particles per unit
area,
MH2
NH2 = (1.1)
µmH2 πR2
has a simple and linear relation with the observed intensity of the CO(1–0) line
through so called CO-to-H2 conversion factor, or XCO :
NH2 = XCO ICO(1−0) (1.2)
So by comparing the two previous equations we can clear a relation between the
molecular gas mass and the intensity of the CO line:
MH2 = XCO µmH2 πR2 ICO(1−0) (1.3)
where mH2 is the mass of one H2 molecule and µ is the correction factor for the
contribution of heavy elements by mass, mainly dominated by the He molecular
weight.
16
The radius R in Equation 1.1 is the radius of the area taken into account for
calculating the column density, when assuming a spherical symmetry in molecular
clouds. In practical terms, as we can see in Equation 1.3, this radius is restricted
to the CO observations (otherwise Equation 1.3 would not make much sense). So
the radius R in Equation 1.3 is the radius in which the CO line intensity ICO was
measured. As we will see in the following chapters, due to the nature of our CO ob-
servations in this particular work, this radius corresponds to the CO(1–0) beamsize
or Full Width at Half Maximum (FWHM).
Now, if we assume a Galactic value for the CO conversion factor, XCO = 2 × 1020
[cm−2 (K km/s)−1 ] = 1.9×1057 [pc−2 (K km/s)−1 ] (Dickman et al. 1986), a correction
factor for heavy elements of 1.36 (based on cosmological abundances, Bolatto et al.
2013), and a mass of the H2 molecule of 1.7 × 10−57 M , a simpler expression for the
molecular mass is obtained:
MH2 [M ] = 4.4πR2 [pc]ICO(1−0) [K km/s] (1.4)
where R is the radius of the CO(1-0) beamsize in parsecs, so it depends on the

distance of the source. This formula for calculating the mass of H2 from the CO line
intensity will be used throughout this entire work.
1.3 Kennicutt-Schmidt relation

From the moment we acknowledge that stars form due to the gravitational col-
lapse of gas inside molecular clouds, a first and direct conclusion can be established:
the star formation must depend on the amount of gas available. And to parameterize
this relation is of key importance to begin to understand the physics of the process
and its evolution throughout time. This is how in nearby galaxies, an empirical
relation has been established between star formation and gas surface densities, the
Kennicutt-Schmidt (K-S) relation, that compares the SFR surface density (ΣSFR )
versus the gas surface density (Σgas ), which, depending on the analysis, can be
molecular gas (H2 ), atomic gas (HI), or both (HI+H2 ). The first work done on this
topic was by Schmidt (1959), setting for the first time the concept of a power-law
relation between the SFR and gas densities. On physical grounds, we may expect
the most fundamental relation to be between the volume densities of star formation
and gas, but because most observations of external galaxies can measure only surface
densities integrated along the line of sight, the most commonly used K-S relation in
terms of densities is:
ΣSFR = AΣN
gas (1.5)
which in log-space translates into a constant slope plot. The coefficient A traces
the absolute SFE (Kennicutt 1998b), and the slope N relates the star formation
17
Figure 1.2: (a) Relationship between the disk-averaged surface densities of star formation and
gas (atomic and molecular) for different classes of star-forming galaxies. Each point represents an
individual galaxy, with the SFRs and gas masses normalized to the radius of the main star-forming
disk. For this plot, a constant XCO factor was applied to all galaxies. The light blue line shows
a fiducial relation with slope N = 1.4 (not intended as a fit to these data). (b) Corresponding
relation between the total (absolute) SFR and the mass of dense molecular gas as traced in HCN.
The dashed gray line is a linear fit, which contrasts with the nonlinear fit in panel a. Figure
adapted from Gao & Solomon (2004).
rate to the gas density present and tells us how close are the two variables from a
linearity. The relation is nearly linear when Σgas considers molecular gas (H2 ), while
is less tight with atomic hydrogen (HI).
Kennicutt (1998a) presented a review of observations of the K-S relation up to the
time, and nearly all that work characterized the relation between the disk-averaged
SFR and gas surface densities in galaxies. Figure 1.2 presents an updated version
of the global K-S in nearby galaxies, taken from Kennicutt & Evans (2012). Each
point is an individual galaxy with the surface density defined as the total gas mass
(molecular plus atomic) or SFR normalized to the radius of the main star-forming
disk, as measured from Hα, Paα, or IR maps.
Most of the galaxies form a tight relation, and the dispersion of the normal
galaxies overall from the average relation ( ± 0.30 dex rms) is considerably higher
than can be attributed to observational uncertainties, which suggests that much
of the dispersion is physical. The form of this integrated K-S relation appears to
be surprisingly insensitive to the SFR environment and parameters such as the
atomic versus molecular fraction. However, some metal-poor galaxies (defined as
Z < 0.3 Z ) deviate systematically from the main relation (Figure 1.2a). These
deviations could arise from a physical change in the star-formation relation, but
more likely reflect a breakdown in the application of a constant XCO factor (Leroy
18
et al. 2011, and references therein). Adopting higher values of XCO for metal-poor
galaxies brings the galaxies much more into accord with the main relation in Figure
1.2a. Observations of low-surface-brightness spiral galaxies by Wyder et al. (2009)
extend the measurements of the integrated star-formation relation to even lower
mean surface densities (Figure 1.2a), where a clear turnover is present. The slope of
the integrated K-S relation is nonlinear, with N ∼ 1.4–1.5 (Kennicutt 1998b) when a
constant XCO factor is applied. This uncertainty range does not include all possible
systematic errors, arising, for example, from changes in XCO or the initial mass
function (IMF) with increasing surface density or SFR. Major systematic changes
in either of these could easily change the derived value of N by as much as 0.2–0.3,
which is what would happen if we would change XCO to a value five times lower in
the dense starburst galaxies: the slope N of the overall K-S relation would increase
from 1.4–1.5 to 1.7–1.9 (Narayanan et al. 2011).
Usually, the K-S relation is parameterized in terms of the total (HI+H2 ) gas
surface density, but one can also explore the dependences of the disk-averaged SFR
densities on the mean HI and H2 surface densities individually. Among galaxies
with relatively low mean surface densities, the SFR density is not particularly well
correlated with either component, though variations in XCO could partly explain the
poor correlation between SFR and derived H2 densities (e.g., Kennicutt 1998b).
The dependence of the SFR on dense molecular gas mass is markedly different,
however. The right panel of Figure 1.2, taken from Gao & Solomon (2004), shows the
relation between the integrated SFRs and the dense molecular gas masses, as derived
from HCN J = 1 → 0 measurements for a sample of normal and starburst galaxies.
In contrast to the correlation with total molecular mass from CO J = 1 → 0, this
relation is linear, implying a strong coupling between the masses of dense molecular
clumps and stars formed, which is largely independent of the galactic star-forming
environment. Wu et al. (2005) have subsequently shown that this linear relation
extends down to the scales of individual star-forming molecular clouds and dense
clumps in the Galaxy. Combined with the MW studies, these dense gas relations for
galaxies suggest that dense clumps are plausible fundamental star-forming units. If
so, the mass fraction of the ISM (and fraction of the total molecular gas) residing
in dense clumps must increase systematically with the SFR.
Finally, the K-S relation gives one more extra piece of information, used through-
out this entire work: the efficiency of the star-forming process. The SFE can be
directly measured as
ΣSFR
SFE = (1.6)
Σgas
where Σgas can consider either just the atomic gas (HI), the molecular gas (H2 )
or the combination of both (HI+H2 ). The inverse of this quantity (i,e. Σgas /ΣSFR )
is called the depletion time τdep , and tell us exactly how long will the process take
to convert all the amount of gas taken under consideration, into stars. For nearby
spiral galaxies these depletion times are often found to be between 2 and 3 Gyrs
19
(Bigiel et al. 2011, Leroy et al. 2008). As we will see in the following chapters, the
efficiency of the star-forming processes in low gas density regions is much lower,
implying longer depletion times.
1.4 Low gas density environments

This thesis focuses on star formation in low gas density regimes, where the typical
gas surface density is less than 10 M pc−2 and the gas is mostly atomic. These
regions can be found in the outer disks of spirals, early-type, low surface brightness
(LSB) and dwarf galaxies, and the characterize for having a very low and sparse
star formation, low metallicities and dust abundances, relatively high shear, low
total gas column densities spread over significant scale heights, a preponderance of
HI compared to H2 , and a comparatively weak stellar potential well (Kennicutt &
Evans 2012).
But even though the cold gas in this regime is predominantly atomic, small and
local concentrations of molecular gas are often found. Star formation is highly dis-
persed, with young clusters and HII regions observed only in regions of unusually
high cold-gas densities. The SFE in these places is very low and seems to be uncor-
related with Σgas , where the SFR has a much larger dynamic range than the local
gas density. Such are the results found in the outer parts of spiral galaxies (Bigiel
et al. 2008, 2010).
Observations in such environments are indeed difficult, due to the weakness of the
emission, but extremely important in the field of star formation. A robust, quanti-
tative picture of how the the environment in the outer disks affects star formation
is crucial if we want to understand the origins of galaxy structure. Many galaxies
sustain a large reservoir of (low column density) gas in their outer disks over evo-
lutionary timescales. To measure the gas consumption timescale in this regime for
many galaxies and compare it to the gas consumption time found for the inner parts
of galaxies provides valuable clues regarding the role of outer disk gas for fueling star
formation over cosmological times (Shlosman et al. 1989, Bauermeister et al. 2010).
Additionally, star formation at large galactocentric radii will affect how chemical
enrichment varies across a galaxy (Gil de Paz et al. 2007b) and plays a critical role
in determining the location and form of the break in the exponential stellar disk
(Pohlen & Trujillo 2006)
But evidences of star formation activity in low gas density mediums are not found
only in the outer parts of galaxies. This thesis also studies regions even farther away
from the central parts of galaxies, where typical star formation occurs, and involves
the intergalactic medium in the Virgo Cluster, where galaxies interact with each
other, affecting their structure and further evolution. The dynamics of these galaxies
are dominated by the gravitational potential of the entire cluster, and thus not only
interact with each other through collisional processes, but also with the cluster as a
whole, and with the intra-cluster medium (ICM), which strips the gas content from
them. This effect is most likely to be the result of ram-pressure, i.e the force that
20
any body, in this case a galaxy, feels when moving through a fluid, like the ICM. The
ram-pressure can strip large amounts of atomic gas from galaxies, up to thousands
of parsecs away, forming large tails of atomic gas, the initial fuel of star formation.
The ICM is a very hostile environment, with temperatures of ∼107 -108 K (Roediger
& Hensler 2005), and destructive X-rays (e.g., Machacek et al. 2004; Fabian et al.
2006; Tamura et al. 2009), and yet surprisingly there has been evidence of molecular
gas clumps, most likely formed in situ, since ram-pressure is not expected to be able
to drag such a dense gas. Furthermore, evidence of recent star formation has also
been observed through Hα emission. Active star formation in such unusual, hostile
and isolated environments makes an ideal laboratory to study star formation under
extreme conditions, and help us to define the limiting conditions for the process
to take place, as well as to disentangle what are the determining factors that can
trigger such an unlikely event.
1.5 Instrumentation
The main data for this thesis were obtained through observations at the IRAM
30 mt single dish radio telescope (Figure 1.3). Located at 2850 meters of altitude in
Pico Veleta in the Spanish Sierra Nevada, this telescope is one of the most sought-
after radio telescopes in the world.
Figure 1.3: IRAM 30-mt single dish radio telescope at Pico Veleta in Sierra Nevada, Spain.
Thanks to its heterodyne Eight MIxer Receiver (EMIR), four double sidebands
are available to cover the millimeter spectral range from 83 to 360 GHz (Carter
et al. 2012). These bands cover the four main atmospheric windows in the millimeter
range, at 3, 2, 1.3 and 0.9 mm (Figure 1.4), and are ideal to study the cold interstellar
matter in its molecular form. Furthermore, these bands are built in a technology
that offers 8 GHz instantaneous bandwidth per sideband and polarization (horizontal
and vertical), and allows three possible combination of dual and simultaneous band
observations.
21
Figure 1.4: Atmospheric transmission between 60 and 400GHz for two precipitable water vapors,
modeled with the ATM model. Additionally the EMIR bands: E090, E150, E230 and E330 (named
after their approximated central frequencies), covering the 3, 2, 1.3 and 0.9 mm atmospheric
windows respectively, and a few important molecular transitions
All the observations done in this thesis were done with the E0/E2 configuration,
which allows simultaneous observations of CO(1–0) and CO(2–1) at 115.271 and
230.538 GHz respectively, in both polarization. The telescope half-power beam
widths at these frequencies are 2200 and 1100 respectively.
Finally, to achieve the spectral resolution needed, both the Wideband Line Mul-
tiple Autocorrelator (WILMA) and the Fast Fourier Transform Spectrometer(FTS)
backends were used, which give an spectral resolution of 2 MHz and 195 kHz re-
spectively. At 115 GHz, these resolutions translate into 5.2 km/s for WILMA and
0.5 km/s for FTS, and at 230 GHz, the corresponding values are 2.6 km/s and 0.25
km/s respectively.
22
Chapter 2
XUV
Contents
2.1 M63 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.1 Observations . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.1.3 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.1.4 Summary and conclusions . . . . . . . . . . . . . . . . . . 48
2.1.5 Publication . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.2 2013 observations . . . . . . . . . . . . . . . . . . . . . . . 63
2.2.1 Data reduction and Results . . . . . . . . . . . . . . . . . 63
2.2.2 Stacking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.3 M83 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.3.1 Instrumentation and Observations . . . . . . . . . . . . . 69
2.3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.3.3 Star formation rate . . . . . . . . . . . . . . . . . . . . . . 72
Ever since the Galaxy Evolution Explorer (GALEX) telescope was launched in
2003 it has mapped numerous galaxies in our local Universe, in both near ultraviolet
(NUV) at 2267Å, and far ultraviolet (FUV) at 1516Å. With its 1.25 degree field of
view, a spatial resolution of 500 , and a sensitivity to stellar populations younger than
a few hundred Myrs, the GALEX telescope is ideal to resolve the spiral structure
of nearby galaxies and map with great detail the ultra-violet (UV) emission coming
from young O and B stars, characterized by ages from a few Myr up to 400 Myr.
During the first years of the mission, the GALEX telescope made a major discovery,
by surprisingly unveiling large disks of UV emission in several nearby spiral galaxies,
that can go up to 3 to 4 times their optical radii (Gil de Paz et al. 2007a). Such
disks of extended ultraviolet (XUV) emission are present in ∼30% of spiral galaxies,
and in many cases go further than the Hα emission, the typical star formation
tracer. The most remarkable examples of XUV disks are present in M63, M83,
NGC2841 and NGC4625 (Thilker et al. 2005, Gil de Paz et al. 2005, 2007a), and
23
they cover a significant fraction of the area detected at 21 cm wavelength, with some
correspondence between the position of the brightest UV complexes and peak of
atomic gas distribution. With these XUV disk of emission, the GALEX telescope has
demonstrated that Hα observations are not always good tracers of star formation,
and can fail to detect a significant population of moderate-age stars in the outermost
disks of spiral galaxies, since Hα traces more recent episodes of star formation. In
these scenarios, UV emission is a much better tracer of star formation, and can be
used to estimate the efficiency of this process.
These XUV regions in the outer parts of spiral galaxies have gained interest
in the past few years in the field of star formation, since the are low metallicity
environments (Henry & Worthey 1999), resembling the conditions of early stages
of spiral galaxies and high-redshift galaxies. These regions are also known to have
low star formation rates (Dong et al. 2008, Bigiel et al. 2010, Alberts et al. 2011).
Moreover, there is growing evidence of cold gas accretion in the local Universe, both
through the arrival and merging of gas-rich satellites and through gas infall from
the intergalactic medium. This new gas could be deposited in the outer regions
of galaxies and form reservoirs for replenishing the inner parts and feeding star
formation (Sancisi et al. 2008), making outer regions good laboratories for scanning
the interface between galaxies and the surrounding intergalactic gas.
The importance of the occurrence of recent and ongoing star formation in the
outer disks of normal spiral galaxies has several important implications. First of all,
it supports the presence of molecular gas in the outskirts of spirals, since stars are
formed within molecular clouds. Furthermore, this suggests the presence of large
reservoirs of hydrogen in the form of H2 , which may contribute to the baryonic
dark matter of spiral galaxies. It also offers the ideal place to study the unresolved
issue of the atomic hydrogen gas origin: is HI mainly a product of the star formation
process; i.e, does it result from the photodissociation of H2 by the UV flux from newly
formed stars (Allen et al. 1986, 2004, Smith et al. 2000), rather than mainly being
a precursor of it? Additionally, the presence of recently formed stellar complexes
at large galactocentric radii also provides a simplified laboratory for determining
the star formation to occur spontaneously (Kennicutt 1989, Martin & Kennicutt
2001). And finally, it allows us to investigate the star formation in quiescent and
low metallicity environments that may affect the star formation density and the
initial mass function.
The aim of the work presented in this chapter is demonstrate the presence of
molecular gas in the outer parts of a few XUV disk galaxies in the nearby universe,
and use it to study the star formation efficiency of the process at extremely low gas
densities, where little is known about the topic.
2.1 M63
For this thesis, the first galaxy studied in the XUV topic was M63 (NGC5055),
by resuming a work started in 2007 with CO observations from the IRAM 30-m
24
telescope. This section presents the results of those observations and the analysis
done in the CO data, in combination with complementary UV and infrared data, to
quantify the SFR and study its efficiency.
M63 is a SA(rs)bc galaxy located at 10.1 Mpc (Leroy et al. 2009), and looks
like a typical spiral galaxy, representative of a large class of local spirals, with no
immediate neighbor, which excludes the potentiality of an object in interaction.
However, M63 is not so ordinary, as it is a remarkable example of a nearby spiral
with a bright and extended XUV disk. In Fig. 2.1 is shown the GALEX NUV and
FUV color-composite image of M63, where an extensive population of UV-bright
star-forming regions and stellar clusters (tracing the O and B stars) is revealed. The
major axis radius of the B-band 25 mag arcsec−2 isophote, the so-called isophotal
or optical radius, is equal to r25 = 35400 = 17.4 kpc in M63. Its NUV and FUV
surface-brightness profiles show a smooth decrease of the XUV emission out to 2.5
times the optical radius (Gil de Paz et al. 2007a).
The 21 cm observations of M63 (see Figure 2.11) also show the presence of a very
large, warped gaseous disk extending out to 40 kpc in the major axis radius (e.g.,
Battaglia et al. 2006). The warp starts around r25 and is exceptionally extended and
symmetric, suggesting a stable dynamics. The measured H I column densities are
larger than 1020 cm−2 up to 70% of the XUV disk. These large H I column densities
plus the extended XUV emission, both suggest the presence of molecular gas out to
large galactocentric radii. Indeed, recent studies have demonstrated that strong H I
emission is on average a good tracer of regions rich in molecular gas (e.g., Crosthwaite
et al. 2002, Nieten et al. 2006), and the XUV emission undeniably betrays the
presence of stars and hence of molecular gas necessary for their formation.
Molecular gas in M63 has been searched in the past in the context of the BIMA/SONG
survey, the first systematic imaging survey of 12 CO(1–0) emission from the centers
and disks of nearby galaxies (Helfer et al. 2003, see their Fig. 43). The reported high-
resolution CO measurements are confined to the very central area of the M63 optical
disk, rgal = 9600 = 0.3 r25 along the major axis, because of the lack of sensitivity of
these data at the border of the map with a primary beam gain drop by a factor of 2.
Single-dish spectra from the FCRAO Extragalactic CO Survey (Young et al. 1995,
see their Fig. 87), despite their moderate sensitivity and low-resolution, show that
the CO emission extends over a larger area than that mapped by BIMA/SONG,
reaching a two times larger galactocentric radius, but still not extending up to the
optical disk limit r25 . The signal measured at rgal = 18000 is equal to ∼ 2 K km s−1 ,
which leaves place for a CO line intensity decrease by a factor of 10 to 20 in the
outermost XUV disk (depending on the CO line width), when aiming at signals as
weak as 10 mK, or weaker, in these extreme regions. More recently, M63 was also
observed by Schruba et al. (2011), but in this work, once again beyond the galac-
tocentric radius rgal = 0.8 r25 , only upper limits are reported for 3 σ detections, still
lacking of high enough sensitivity. Finally, Leroy et al. (2009) mapped the CO(2–1)
line in M63 over the full optical disk, as part as the HERACLES survey on the
IRAM 30 m telescope, and obtained reliable measurements out to rgal = 0.68 r25 .
The aim of the present work was to go further than all these observations, and
25
achieve a good sensitivity beyond the r25 limit, to confirm the presence of H2 sug-
gested by the extended HI and FUV-NUV emission.
42◦ 10′
42◦ 05′
DEC. (J2000)
42◦ 00′
41◦ 55′
30s 13h 16m 00s 30s 13h 15m 00s

R.A. (J2000)
Figure 2.1: False-color RGB image of M63. The image was produced using the arcsinh function,
which allows to show faint structures while simultaneously preserving brighter structures in the
field, such as the spiral arms of large galaxies (Lupton et al. 2004). The GALEX NUV and FUV
emissions are plotted in red and blue, respectively, and the composition of NUV (80%) and FUV
(20%) is plotted in green. The green ellipse represents the B-band 25 mag arcsec−2 isophote limit,
r25 , or optical limit. The red solid line shows the CO mapping we performed along the major axis
of M63 from the center out to rgal = 57200 = 1.6 r25 . The red box encloses the chosen bright UV
region, located at rgal = 48300 = 1.36 r25 .
2.1.1 Observations
The observations were performed with the IRAM 30-m telescope at Pico Veleta,
Spain, during a first run on September 10–17, 2007 under poor weather conditions,
and during a second run on November 2, 16 and 27 and December 1, 2007 under
good-to-excellent weather conditions.
Simultaneous CO(1–0) and CO(2–1) observations were done with the now de-
commissioned single pixel heterodyne receivers, with two of them centered on the
26
12
CO(1–0) line at 115.271 GHz, and two on the 12 CO(2–1) line at 230.538 GHz.
The telescope half-power beam widths at these two frequencies are 2200 and 1100 , re-
spectively. The data were recorded using the VESPA autocorrelator with 640 MHz
bandwidth and 1.25 MHz resolution at 3 mm, and two 1 MHz filter banks (512 chan-
nels each) at 1 mm. The resulting velocity coverage at 115.271 GHz is 1665 km s−1
with a spectral resolution of 3.2 km s−1 . The corresponding values at 230.538 GHz
are 666 km s−1 and 1.3 km s−1 .
The observing strategy for this work was of key importance to achieve the nec-
essary sensitivity. For this, an arrange of single ON+OFF positions was created,
instead of doing a raster mapping, which requires much more integration time to
reach a good S/N. First, a CO mapping was performed along the major axis of
the M63 disk from the center of the galaxy located at (J2000) RA = 13h 15m 49.3s ,
Dec = +42◦ 01m 45.4s out to the galactocentric radius rgal = 57200 = 1.6 r25 by
aligning 27 single pointings with a spatial sampling of 2200 in the radial direction
sustaining a position angle PA = 105◦ . This alignment of pointings is shown in
Figure 2.2. Second, a bright UV region was mapped in the outer regions of the M63
disk centered on (J2000) RA = 13h 15m 07.0s , Dec = +42◦ 00m 00.0s and located at
the galactocentric radius rgal = 48300 with 6×2 = 12 pointings following the sequence
from (−6600 ; 000 ) to (+4400 ; +2200 ) offsets relative to the central coordinates with a
2200 spatial sampling in the right ascension direction and same in the declination
direction. In Figure 2.2 we can see the positions of these 12 pointings in the UV
region. Observations were performed in wobbler switching mode with the maximum
symmetrical azimuthal wobbler throw of 24000 allowed, corresponding, for this spiral
galaxy located at a distance of 10.1 Mpc, to 11.8 kpc in projected distance.
2.1.2 Results
The data were reduced with the CLASS software from the GILDAS package.
All the spectra obtained with the two receivers tuned on the 12 CO(1–0) line and
corresponding to scans at the same pointings were summed up without any smooth-
ing. On the other hand, the spectra obtained with the two receivers tuned on the
12
CO(2–1) line were first Hanning smoothed to a resolution of 2.6 km s−1 , since the
expected average CO line width is of the order of 10 km s−1 , before being summed
up when corresponding to scans at the same pointings. No baseline subtraction was
performed on individual spectra before summing, as we simply did a linear sum. In
Tables 2.1 and 2.2 (columns 4 and 7) we list the achieved 1 σ rms in mK at 115
GHz (or 3 mm) and 230 GHz (1 mm) for each pointing of the M63 disk map and
the UV region map, respectively1 . These rms values were obtained with a baseline
subtraction of degree 0, and with windows set to 300 − 400 km s−1 , both at 3 mm
and 1 mm, as defined by the velocity positions of detected CO(1–0) and CO(2–1)
1
The efficiencies of the ABCD receivers on the IRAM 30 m telescope are
5.9 Jy K−1 at 3 mm and 7.2 Jy K−1 at 1 mm (see the online IRAM wiki pages at
http://www.iram.es/IRAMES/mainWiki/Iram30mEfficiencies).
27
Figure 2.2: Positions of observing pointings to map the radial axis of M63, as well as its bright
UV region, overlaid in the same NUV-FUV composite image of Figure 2.18. Mapping CO through
individual beam pointings, has proben to be a very efficient technique to obtain deep observations
of specific, not extended areas of a galaxy. To better visualize the spatial resolution of our data,
pointings circles have been drawn with a 1100 width, same as the CO(2–1) FWHM.
28
lines2 . At 3 mm we obtained on average rms values between 4 and 21 mK in the
M63 disk map, while for seven pointings around the isophotal radius, r25 , and for
all the pointings of the UV region map we pushed the rms limit down to 4 to 7 mK.
The results of our CO emission mapping of M63 along the major axis of its disk
and over the bright UV region at rgal = 1.36 r25 are presented in Tables 2.1 and
2.2, respectively. We provide the CO(1–0) and CO(2–1) line full widths at half
maximum, FWHM (columns 5 and 8) and the integrated CO(1–0) and CO(2–1)
line fluxes (columns 6 and 9) at each pointing of the mapping, as determined from
fitting Gaussian functions to the CO profiles obtained by summing up all spectra
corresponding to all scans per pointing. The CO(1–0) and CO(2–1) line FWHM
are in the range from 9 to 53 km s−1 and from 6 to 39 km s−1 , respectively. They
are greater than those expected for individual giant molecular clouds (GMCs) that
have typical line widths of ∼ 10.4 km s−1 (Solomon et al. 1987), except for a few
pointings around r25 and over the UV region. This implies that mostly an ensemble
of molecular clouds is emitting per beam of ∼ 1 and 0.5 kpc at 3 and 1 mm,
respectively.
In Table 2.1 we also list the molecular gas masses per pointing (column 11),
through Eq. 1.4, using the “standard” Galactic CO(1–0)-to-H2 conversion factor
XCO = 2 × 1020 cm−2 (K km s−1 )−1 (Dickman et al. 1986) and including a correction
for helium. The inferred masses are in the range MH2 = (1.5 − 322) × 106 M .
Figure 2.3 shows the summed-up spectra of all scans per pointing of the M63 disk
map along the major axis, while Fig. 2.4 shows the resulting spectrum obtained by
summing up all spectra of all scans at the 12 pointings used to map the UV region.
The radial mapping (Fig. 2.3) clearly shows a detection of the CO(1–0) emission
out to the galactocentric radius rgal = 35200 , namely out to the isophotal radius r25 =
35400 = 17.4 kpc defined as the radius of the B-band 25 mag arcsec−2 isophote. A
severe drop of the CO flux is nevertheless observed as a function of the galactocentric
radius. Beyond r25 and out to 1.62 r25 , the limit of our CO search along the M63
major axis, no CO emission is detected anymore (Table 2.1). Even by summing
up all the spectra corresponding to all scans at the 10 outermost pointings and by
smoothing the resulting spectrum to a resolution of 9.7 km s−1 , corresponding to
the typical FWHM of the CO(1–0) lines detected at the outermost pointings, no
CO emission is observed. We do, however, derive a stringent 3 σ upper limit on
the integrated CO(1–0) line flux at rgal > r25 of FCO(1−0) < 0.06 K km s−1 , when
assuming a line width of 10 km s−1 . This is equivalent to a molecular gas mass of
MH2 < 3.9 × 105 M . On the other hand, the CO(2–1) emission is securely detected
out to rgal = 24200 only, i.e. 0.68 r25 , in agreement with the CO(2–1) mapping by
Leroy et al. (2009), while there are two very tentative detections at 2 σ at the
pointings rgal = 0.93 r25 and r25 (Table 2.1). The CO(2–1) line thus appears to be
excited over about 2/3 of the optical disk. The corresponding CO luminosity ratios,
r2,1 = L0CO(2−1) /L0CO(1−0) , can be found in Table 2.1 (Col. 10). They varie randomly
2
Except for the very few inner pointings at rgal = 0, 2200 , and 4400 for which a larger window
was set, given the larger widths of the CO(1–0) and CO(2–1) lines.
29
Table 2.1. CO mapping along the major axis of the M63 disk.
CO(1–0) CO(2–1)
rgal rgal Time rmsa FWHMb Fc rmsa FWHMb Fc r2,1 d MH2 e ΣH2 f
[00 ] [r25 ] [min] [mK] [km s−1 ] [K km s−1 ] [mK] [km s−1 ] [K km s−1 ] [106 M ] [M pc2 ]
0 0 31 11. 200 49.8 ± 2.2 16. 192 39.8 ± 3.1 0.24 321.8 ± 14.2 125.7 ± 5.6
22 0.06 8 18. 53 24.5 ± 0.9 28. 27 12.6 ± 0.8 0.16 158.3 ± 5.8 61.8 ± 2.3
44 0.12 8 19. 41 16.4 ± 0.8 33. 21 9.0 ± 0.7 0.17 106.0 ± 5.2 41.4 ± 2.0
66 0.19 8 19. 23 9.8 ± 0.4 32. 16 4.2 ± 0.5 0.13 63.3 ± 2.6 24.7 ± 1.0
88 0.25 8 18. 20 6.3 ± 0.4 32. 20 3.4 ± 0.6 0.17 40.7 ± 2.6 15.9 ± 1.0
110 0.31 8 19. 26 12.6 ± 0.5 30. 21 8.2 ± 0.6 0.20 81.4 ± 3.2 31.8 ± 1.3
132 0.37 8 21. 22 5.1 ± 0.5 35. 20 2.6 ± 0.7 0.16 33.0 ± 3.2 12.9 ± 1.3
154 0.44 23 13. 29 5.3 ± 0.4 22. 39 2.5 ± 0.9‡ 0.15 34.3 ± 2.6 13.4 ± 1.0
176 0.50 23 13. 19 5.4 ± 0.2 24. 27 2.8 ± 0.6 0.16 34.9 ± 1.3 13.6 ± 0.5
198 0.56 23 13. 21 1.8 ± 0.3 22. 18 0.82 ± 0.40‡ 0.14 11.6 ± 1.9 4.5 ± 0.8
220 0.62 31 11. 16 0.98 ± 0.17 11.† 10 < 0.33 < 0.11 6.3 ± 1.1 2.5 ± 0.4
242 0.68 101 6. 11 0.30 ± 0.06 8. 6 0.17 ± 0.05 0.17 1.9 ± 0.4 0.76 ± 0.15
264 0.75 101 6. 14 0.45 ± 0.08 4.† 10 < 0.12 < 0.08 2.9 ± 0.5 1.1 ± 0.2
286 0.81 55 10. 21 0.88 ± 0.21 9.† 10 < 0.27 < 0.10 5.7 ± 1.4 2.2 ± 0.5
308 0.87 47 11. 14 0.81 ± 0.15 11.† 10 < 0.33 < 0.13 5.2 ± 1.0 2.0 ± 0.4
330 0.93 63 7. 22 0.85 ± 0.15 9. 34 0.38 ± 0.31‡ 0.14 5.5 ± 1.0 2.1 ± 0.4
352 0.99 155 5. 9 0.23 ± 0.04 6. 34 0.29 ± 0.21‡ 0.38 1.5 ± 0.3 0.58 ± 0.10
30
374 1.06 109 4.† 10 < 0.12 4.† 10 < 0.12 ··· ··· ···
396 1.12 117 4.† 10 < 0.12 3.† 10 < 0.09 ··· ··· ···
418 1.18 93 5.† 10 < 0.15 5.† 10 < 0.15 ··· ··· ···
··· ··· ··· ··· ··· ··· ··· ··· ··· ··· ···
473 1.34 381 2.† 10 < 0.06 2.† 10 < 0.06 ··· < 0.39 < 0.15
Note. — The radius of the B-band 25 mag arcsec−2 isophote, the so-called isophotal radius, is equal to r25 = 35400 = 17.4 kpc
in M63. It is often used as a reference to express the relative galactocentric radius, rgal . The last separate line of the Table gives
the values obtained when summing all scans at the 10 outermost pointings with no CO detection from 1.06 r25 to 1.62 r25 and by
smoothing the resulting 3 mm and 1 mm spectra to a resolution of 9.7 and 10.4 km s−1 , respectively.
a Rms noises at, respectively, 3 and 1 mm, measured per channel of 3.2 and 1.3 km s−1 . The values marked with † correspond to
rms noises measured per smoothed channel of 9.7 and 10.4 km s−1 , respectively.
b Full widths at half maximum of, respectively, CO(1–0) and CO(2–1) lines as determined from fitting Gaussian profiles.
c Integrated
CO(1–0) and CO(2-1) line intensities as determined from fitting Gaussian profiles. The values marked with ‡ correspond
to potentially less reliable measurements, because of a CO line detection at 2 σ only. Upper limits are 3 σ and are calculated assuming
a 10 km s−1 line width.
d CO
luminosity correction factor defined as r2,1 = L0CO(2−1) /L0CO(1−0) , needed to correct the lower Rayleigh-Jeans brightness
temperature of the 2–1 transition relative to 1–0.
e H mass calculated from the CO(1–0) luminosity and by adopting the “Galactic” CO-to-H conversion factor, X
2 2 CO = 2 ×
1020 cm−2 (K km s−1 )−1 (Dickman et al. 1986). The applied formula is: MH2 (M ) = 4.4L0CO(1−0) (K km s−1 pc2 ), where a factor
of 1.36 is included to account for helium.
f H surface density calculated from the integrated CO(1–0) line intensity and by adopting the CO-to-H conversion factor given
2 2
under (e) . The applied formula is: ΣH2 (M pc−2 ) = 4.4 cos(i)ICO(1−0) (K km s−1 ), where i = 55◦ is the inclination of M63 (Leroy
et al. 2008).
Table 2.2. CO mapping over the bright UV region at 1.36 r25 .
CO(1–0) CO(2–1)
Offsets Time rmsa FWHMb Fc rmsa FWHMb Fc r2,1 d MH 2 e ΣH2 f
00
( ) (00 ) (min) (mK) (km s−1 ) (K km s−1 ) (mK) (km s−1 ) (K km s−1 ) (106 M ) (M pc2 )
−66 0 118 5 10 < 0.15 6 10 < 0.18 ··· < 0.97 < 0.38
−66 +22 117 5 11 0.14 ± 0.06‡ 6 10 < 0.18 < 0.39 0.90 ± 0.39 0.35 ± 0.15
−44 0 186 4 10 < 0.12 4 10 < 0.12 ··· < 0.78 0.30
−44 +22 164 4 10 < 0.12 6 10 < 0.18 ··· < 0.78 0.30
−22 0 171 4 14 0.21 ± 0.06 5 10 < 0.15 < 0.22 1.36 ± 0.39 0.53 ± 0.15
−22 +22 148 4 10 < 0.12 6 10 < 0.18 ··· < 0.78 < 0.30
0 0 164 4 10 < 0.12 5 10 < 0.15 ··· < 0.78 < 0.30
0 +22 148 5 10 < 0.15 6 10 < 0.18 ··· < 0.97 < 0.38
+22 0 123 5 10 < 0.15 7 10 < 0.21 ··· < 0.97 < 0.38
+22 +22 116 5 10 < 0.15 8 10 < 0.24 < 0.97 < 0.38
31
···
+44 0 63 7 10 < 0.21 9 10 < 0.27 ··· < 1.36 < 0.53
+44 +22 62 7 10 < 0.21 8 10 < 0.24 ··· < 1.36 < 0.53
all offsets 1932 1 23 0.11 ± 0.02 2 10 < 0.06 < 0.17 0.71 ± 0.13 0.28 ± 0.05
Note. — The last separate line gives the values obtained when summing all scans at the 12 offset pointings.
a Rms noises at, respectively, 3 and 1 mm, measured per channel of 3.2 and 1.3 km s−1 .
b Full widths at half maximum of, respectively, CO(1–0) and CO(2–1) lines as determined from fitting Gaussian profiles.
c Integrated CO(1–0) and CO(2–1) line intensities as determined from fitting Gaussian profiles. The values marked with ‡ correspond
to potentially less reliable measurements, because of a CO line detection at 2 σ only. Upper limits are 3 σ and are calculated assuming a
10 km s−1 line width.
d,e,f CO luminosity correction factor, H2 mass, and H2 surface density, respectively, calculated using the same prescriptions as in
Table 2.1.
Figure 2.3: Mosaic of CO spectra obtained when mapping the M63 disk along its major axis
from the center of the galaxy out to the isophotal radius, r25 . The CO(2–1) spectra (red thin
line) are overplotted on the CO(1–0) spectra (black thick line). While the CO(1–0) emission is
clearly detected out the optical radius, the CO(2–1) appears to be confined to rgal . 0.68 r25 . The
bottom-right panel shows the CO(1–0) and CO(2–1) spectra obtained when summing all scans at
the 10 outermost pointings from rgal = 1.06 r25 to 1.62 r25 and smoothed to a resolution of 9.7 and
10.4 km s−1 , respectively. No CO emission is detected at these outermost pointings.
32
Figure 2.4: CO(1–0) spectrum obtained by summing up all spectra at the 12 pointings done to
map the bright UV region at rgal = 1.36 r25 . There is a clear CO emission detection in this region
of M63 well beyond the optical radius. CO(2–1) remains undetected.
between 0.13 and 0.24 (in the center of M63) along the galactocentric radius, without
showing any radial evolution. The absence of a clear detection of the CO(2–1) line
beyond rgal = 0.68 r25 suggests that the subthermal excitation sets in around this
galactocentric radius. However, this has to be considered with caution, because the
major axis pointings probe only a small portion of the M63 disk at each radius and
the Leroy et al. (2009) CO(2–1) map of the entire M63 disk is not very deep, and
because there seems still to be plenty of star forming regions at rgal > 0.68 r25 with
high enough densities (locally) to thermalize the CO(2–1) line (see Fig. 2.1).
Whereas our survey for CO emission along the major axis of the M63 disk seems
to indicate that the molecular gas is confined to the isophotal radius r25 of the
galaxy, we find a convincing CO(1–0) detection in the selected bright UV region
at rgal = 1.36 r25 . The sum of all spectra corresponding to all scans at the 12
pointings done to map the UV region leads to a CO(1–0) detection at 5.5 σ with an
integrated line flux FCO(1−0) = 0.11 ± 0.02 K km s−1 (Fig. 2.4). This corresponds to
a molecular gas mass of MH2 = 7.1 × 107 M . The CO(1–0) emission appears to be
maximal at the offset (−2200 ; 000 ), where it is detected at 3 − 4 σ (Table 2.2). The
CO(2–1) emission is, on the other hand, not detected down to an integrated line flux
FCO(2−1) < 0.06 K km s−1 , obtained when summing up all the spectra corresponding
to all scans at the 12 pointings.
2.1.3 Analysis
A few hundreds of nearby galaxies have been the object of intensive surveys
for molecular gas (e.g., Young & Scoville 1991, Young et al. 1995, Helfer et al.
2003, Leroy et al. 2009). However, for very few only, deep CO searches in their
outer disk regions have been undertaken. Here, we show the existence of molecular
gas up to the isophotal radius, r25 , and beyond in the nearby spiral galaxy M63.
This is the fourth of such a detection of molecular gas in the outskirts after the
spiral galaxies NGC 4414 (Braine & Herpin 2004), NGC 6946 (Braine et al. 2007),
and M33 (Braine et al. 2010). To analyse the impact of our observations in the
33
Table 2.3. Public ancillary data for M63.
Band or line Telescope FWHM Reference
[arcsec]
CO(2–1) IRAM 13.4 Leroy et al. (2009)

FUV GALEX 4.3 Gil de Paz et al. (2007a)
NUV GALEX 5.3 Gil de Paz et al. (2007a)
Hα KPNO 0.38 Kennicutt et al. (2008)
24 µm Spitzer 6 Dale et al. (2009)
H I (21 cm) VLA 6 Walter et al. (2008)
context of star formation in the outer disk of M63, we present the radial profiles
of our CO observations along with complementary data of other star formation
tracers (FUV, NUV, Hα, 24 µm) and H I, and we investigate the behaviour of
the Kennicutt-Schmidt relation across the galaxy and beyond r25 in the bright UV
region. Characterizing the Kennicutt-Schmidt relation in the outer disk regions,
namely in environments with low metallicities, gas temperatures, excitations, and
gas densities, all properties hostile to star formation, is of our main interest.
Radial profiles
In order to compare the CO distribution in M63 along the major axis of its disk
to H I and other star formation tracers’ distributions, we built comparative radial
profiles of the CO, H I, FUV, NUV, Hα, and 24 µm emission. All together, they
provide a complete view of the past and on-going star formation that took or takes
place in the galaxy and of the ingredients needed to sustain the star formation:
molecular and atomic gas. The respective data are taken from the literature and
are described in Table 2.3. All these images are presented in Figure 2.5, where it can
be appreciated the size of the area mapped in each one of them, as well as the extent
of the emission of each tracer. A close look at the bright UV region at rgal = 1.36 r25
is shown in Fig. 2.6, through multiwavelength panels, from left to right, of the NUV,
24 µm, and H I 21 cm emission.
As these complementary data were obtained with different instruments and hence
at various FWHM resolutions, we first convolved all these data to the resolution of
our CO(1–0) data which is equal to the half-power beam width of the IRAM 30 m
telescope at 3 mm, i.e. 2200 . The images were then rotated by 20◦ in a north-to-
west direction to have the major axis in the horizontal axis, and deprojected by
correcting for a 55◦ inclination angle to bring the galaxy face-on. As an example,
the corresponding convolved, rotated and deprojected images of M63 for 24µm,
FUV and HI are shown in Figure 2.7. Finally, we made aperture photometry on
these processed images inside apertures of 2200 diameter using the ‘qphot’ task of
the ‘digiphot.apphot’ package of IRAF3 , and we subtracted an average sky value
3
IRAF is distributed by the National Optical Astronomy Observatories, which are operated by
the Association of Universities for Research in Astronomy, Inc., under cooperative agreement with
the national Science Foundation.
34
Ha HI
24mu NUV
Ha Ha
CO(2-1) HERACLES
Figure 2.5: Maps of complemetary data for M63, taken from the literature (Table 2.3). The r25
or optical limit is drawn as a green ellipse, as well as the 27 single pointings of our observations
along the radial axis of the galaxy, and the 12 pointings for the UV region (in the same way as in
Figure 2.2).
35
Figure 2.6: Multiwavelength views of the bright UV region at rgal = 1.36 r25 , showing from left
to right the NUV, 24 µm, and H I 21 cm emission in flux units of 10−9 MJy sr−1 , MJy sr−1 , and
K km s−1 , respectively, as indicated by the color bars. The circles of 2200 diameter refer to the
positions of the 12 pointings used to map the CO(1–0) emission. The Hα emission is undetected
in this external region of the M63 galaxy.
measured away from the galaxy. The photometric measurements were performed
inside 2200 width apertures (same as our CO(1–0) data FWHM) at the 27 positions
of the radial cut along the M63 major axis used for the CO radial mapping, as well
as at the 12 positions used to map CO in the external bright UV region, following
the same spacing and pointings exactly as in our observations . These photometric
results are presented in Table 2.4 for the pointings a long the radial axis, and in Table
2.5 for the UV region, including our CO(1–0) observations and the complementary
data described in Table 2.3.
Using the results of Tables 2.4 and 2.5, we created radial profiles for the CO(1–0),
CO(2–1), FUV, NUV, H I, Hα, and 24 µm emission, including both, the radial axis
and the UV region. This is presented in Fig. 2.8 (connected data points), where
the measured fluxes are plotted in a normalized flux scale in order to allow relative
comparisons. The dashed vertical line represents the r25 limit. All the profiles
nicely follow the spiral structure observed at galactocentric radii rgal ' 130, 200,
and probably 26000 . They all show a severe drop with the galactocentric radius,
except the profile of the atomic gas which has a completely different behaviour.
We observe that close to the r25 limit the star formation tracers and CO begin to
considerably vanish and, beyond the r25 limit, they all are practically absent with
the exception of the faint XUV emission. The H I emission is very strong beyond the
r25 limit. This shows the importance of looking for molecular gas beyond the optical
limit, where evidence of star formation is clearly brought by the UV emission.
While the CO emission is not detected beyond the r25 limit in the M63 major
axis map, it is again detected in the external UV region at rgal = 1.36 r25 . The
corresponding CO(1–0) fluxes, multiplied by a factor of 40 to make them visible
(this shows the deepness of our data), with H I and other star formation tracer
fluxes as measured in the UV region are also shown in Fig. 2.8 (dashed box). The
FUV and NUV emissions are observed in all pointings over the external UV region,
while the Hα emission is undetected and a faint 24 µm emission, whose fluxes are
multiplied by a factor of 200 to make them visible, is observed in only 8 pointings
36
Figure 2.7: Convolved, rotated and deprojected images of M63 corresponding to the 24µm, FUV
and HI emission. Our observing CO pointings are overlaid in the same way as in Figure 2.2.
37
Table 2.4. Photometry results obtained for the complementary data of M63 in its
radial cut.
rgal rgal CO(1–0) † CO(2–1) ‡ FUV NUV HI Hα 24µm ΣSFR (FUV) ΣSFR (24µm) ΣHI ΣH2
[arcsec] [kpc] [K km/s] [K km/s] [MJy/sr] [MJy/sr] [K km/s] [erg/sec] [MJy/sr] [M yr−1 kpc−2 ] [M yr−1 kpc−2 ] [M pc−2 ] [M pc−2 ]
1 0 0.0 49.8 79.3 2.8E-08 3.6E-08 230.5 2.2E-10 33.5 2.3E-09 1.07E-01 4.6 125.7
2 22 1.1 24.5 46.0 3.7E-08 4.3E-08 267.5 2.2E-10 28.6 3.0E-09 9.2E-02 5.3 61.8
3 44 2.2 16.4 42.1 3.6E-08 3.6E-08 255.4 1.6E-10 18.8 2.9E-09 6.0E-02 5.1 41.4
4 66 3.2 9.8 16.6 2.6E-08 2.5E-08 247.9 1.1E-10 12.9 2.2E-09 4.1E-02 5.0 24.7
5 88 4.3 6.3 15.7 1.6E-08 1.4E-08 183.9 3.7E-11 6.7 1.3E-09 2.1E-02 3.7 15.9
6 110 5.4 12.6 10.9 1.0E-08 9.0E-09 249.7 2.9E-11 5.7 8.4E-10 1.8E-02 5.0 31.8
7 132 6.5 5.1 11.9 1.6E-08 1.1E-08 343.8 6.1E-11 8.9 1.3E-09 2.9E-02 6.9 12.9
8 154 7.5 5.3 11.1 1.1E-08 8.3E-09 288.8 2.4E-11 4.0 9.3E-10 1.3E-02 5.8 13.4
9 176 8.6 5.4 8.9 8.4E-09 5.5E-09 313.1 2.8E-11 3.6 6.8E-10 1.2E-02 6.3 13.6
10 198 9.7 1.8 9.5 1.2E-08 6.6E-09 358.9 2.9E-11 4.2 9.4E-10 1.3E-02 7.2 4.5
11 220 10.8 1.0 4.6 4.2E-09 3.7E-09 261.3 7.3E-12 1.4 3.4E-10 4.5E-03 5.2 2.5
12 242 11.9 0.3 0.7 1.8E-09 4.5E-09 177.8 2.3E-11 0.6 1.5E-10 1.8E-03 3.6 0.8
13 264 12.9 0.5 0.2 1.6E-09 5.9E-09 153.2 2.0E-11 0.3 1.3E-10 1.1E-03 3.1 1.1
38
14 286 14.0 0.9 0.6 2.0E-09 3.4E-09 179.4 7.7E-12 0.2 1.7E-10 7.2E-04 3.6 2.2
15 308 15.1 0.8 0.9 5.2E-09 3.0E-09 220.3 8.6E-12 0.3 4.3E-10 1.1E-03 4.4 2.0
16 330 16.2 0.9 1.4 4.9E-09 2.7E-09 283.6 4.1E-12 0.3 4.0E-10 1.0E-03 5.7 2.2
17 352 17.2 0.2 0.7 2.7E-09 1.8E-09 433.2 2.3E-12 0.4 2.2E-10 1.2E-03 8.7 0.6
18 374 18.3 0.0 0.4 1.9E-09 1.1E-09 423.8 3.1E-12 0.3 1.6E-10 8.6E-04 8.5 0.0
19 396 19.4 0.0 0.6 1.1E-09 6.1E-10 312.5 7.1E-13 0.1 9.1E-11 2.4E-04 6.3 0.0
20 418 20.5 0.0 0.5 1.1E-09 5.6E-10 191.9 2.5E-12 0.0 8.6E-11 5.1E-05 3.8 0.0
21 440 21.5 0.0 0.0 5.9E-10 3.4E-10 67.8 4.4E-12 0.0 4.7E-11 0.0 1.4 0.0
22 462 22.6 0.0 0.0 9.0E-11 5.0E-11 5.9 ··· 0.0 7.3E-12 0.0 0.1 0.0
23 484 23.7 0.0 0.0 1.4E-10 -4.2E-12 0.0 ··· 0.0 1.1E-11 0.0 0.0 0.0
24 506 24.8 0.0 0.0 1.4E-10 2.8E-11 0.0 ··· 0.0 1.1E-11 0.0 0.0 0.0
25 528 25.9 0.0 0.0 0.0 -1.3E-11 0.0 ··· 0.0 0.0 0.0 0.0 0.0
26 550 26.9 0.0 0.0 2.3E-11 -6.1E-11 0.0 ··· 0.0 1.9E-12 0.0 0.0 0.0
27 572 28.0 0.0 0.0 4.0E-11 -7.5E-11 0.0 ··· 0.0 3.3E-12 0.0 0.0 0.0
Note. — † CO(1–0) data from this work. ‡CO(2–1) data from HERACLES (Leroy et al. 2009). The Hα map from (Kennicutt et al. 2008) does not fully cover the mapping along
the radial axis (see Figure 2.5).
Table 2.5. Photometry results obtained for the complementary data of M63 in its
UV region.
offsets CO(1–0) † CO(2–1) ‡ FUV NUV HI Hα 24µm ΣSFR (FUV) ΣSFR (24µm) ΣHI ΣH2
[arcsec] [arcsec] [K km/s] [K km/s] [MJy/sr] [MJy/sr] [K km/s] [erg/sec] [MJy/sr] [M yr−1 kpc−2 ] [M yr−1 kpc−2 ] [M pc−2 ] [M pc−2 ]
1 0 22 <0.15 0 2.14E-09 7.40E-10 170.0 ··· 0.011 1.74E-10 3.56E-05 3.40 <0.38
2 0 0 <0.12 0 3.60E-09 1.76E-09 180.4 ··· 0.030 2.91E-10 9.57E-05 3.61 <0.30
3 -22 0 0.21 0 3.57E-09 1.76E-09 237.8 ··· 0.069 2.89E-10 2.20E-04 4.76 0.53
4 -22 22 <0.12 0 2.21E-09 7.16E-10 198.5 ··· 0.039 1.79E-10 1.25E-04 3.97 <0.30
5 22 0 <0.15 0 3.30E-09 1.59E-09 143.2 ··· 0.003 2.67E-10 1.07E-05 2.86 <0.38
39
6 -44 22 <0.12 0 2.00E-09 8.64E-10 179.2 ··· 0.041 1.62E-10 1.33E-04 3.59 <0.30
7 22 22 <0.15 0 2.63E-09 8.04E-10 153.1 ··· 0.000 2.13E-10 0.00 3.06 <0.38
8 44 0 <0.21 0 2.54E-09 1.21E-09 113.7 ··· 0.000 2.06E-10 0.00 2.27 <0.53
9 44 22 <0.21 0 1.75E-09 4.58E-10 111.5 ··· 0.000 1.42E-10 0.00 2.23 <0.53
10 -44 0 <0.12 0 1.90E-09 8.88E-10 180.3 ··· 0.018 1.54E-10 5.84E-05 3.61 <0.30
11 -66 22 0.14 0 1.21E-09 6.44E-10 123.4 ··· 0.020 9.77E-11 6.34E-05 2.47 0.35
12 -66 0 <0.15 0 7.22E-10 2.18E-10 88.8 ··· 0.000 5.84E-11 0.00 1.78 <0.38
Note. — † CO(1–0) data from this work. ‡CO(2–1) data from HERACLES (Leroy et al. 2009). Offsets are relative to position number 2, marked as a red circle in the UV region in
each panel of Figure 2.5. The Hα map from (Kennicutt et al. 2008) does not cover the UV region (see Figure 2.5).
rgal [kpc]
0 5 10 15 20 25 30
CO(2-1) (HERACLES)
1.0
FUV
NUV
HI
0.8
Hα
24µm
UV reg CO(1-0) (this work)
Normalized flux
0.6
0.4
0.2
CO(1-0) x 40 in UV reg
24µm x 200 in UV reg
0.0
0 100 200 300 400 500 600 700

rgal [arcsec]
Figure 2.8: Radial profiles of FUV, NUV, Hα, 24 µm, different star formation tracers, and H I
taken from the literature (see Table 2.3), as measured along the major axis of the M63 disk, and
which we compare to our CO(1–0) radial profile (red stars) and to the CO(2–1) radial profile from
the HERACLES survey (Leroy et al. 2009). The measured fluxes are plotted in a normalized
flux scale in order to allow relative comparisons. All the star formation tracers nicely follow the
spiral arm structure at rgal ' 130, 200, and probably 26000 . We observe a severe drop with the
galactocentric radius for all star formation tracers, as well as for CO, such that close to the r25
limit (dashed line) and beyond all tracers and CO are practically absent with the exception of XUV
and H I. The CO emission is again detected in the external bright UV region. The measurements
obtained in this UV region are shown in the dashed box at the correct galactocentric radii. We
multiply the CO(1–0) fluxes by a factor of 40 to make them visible. The double data points plotted
at each rgal correspond to the parallel pointings used to map the UV region (see Table 2.2). This
external UV region has relatively high FUV and NUV emissions and is dominated by H I.
40
(see also Fig. 2.6). It is interesting to point out that the FUV and NUV fluxes in
the external UV region are similar to those at the pointing at r25 , the galactocentric
limit where CO is still detected along the M63 major axis. Beyond r25 the FUV and
NUV fluxes decrease and get weaker than in the external UV region. This suggests a
tight correlation between the XUV flux, namely the intensity of star formation, and
the CO flux which should trace the amount of molecular gas, since the CO emission
starts to be detected at a given XUV flux level. If correct, the CO emission is
not detected beyond r25 along the major axis mapping, because the amount of CO,
corresponding to the molecular gas needed to sustain the star formation which is
taking place there, falls below our detection threshold. Consequently, it is likely
that molecular gas is present in the outer regions of the M63 disk, at least as long
as the XUV emission is present.
The considerably larger amount of H I observed in the external UV region relative
to CO leads to speculate that H I is more likely a precursor of H2 rather than a
product of photo-dissociation due to UV radiation. However, the situation might be
more complicated in reality. The main complication we may invoke is the reliability
of CO as a proxy of H2 . As previously discussed, and also shown in Fig. 2.8,
the absence of a clear CO(2–1) line detection beyond rgal = 0.68 r25 may betray a
decrease in the excitation temperature with the galactocentric radius. If this is the
case, it may be that even if substantial H2 is present in the outer regions of the M63
disk, the CO lines may be quite weak, hence remain undetected.
Star formation rate, efficiency, and Kennicutt-Schmidt relation
Our acquired CO measurements offer the opportunity to analyse the Kennicutt-

Schmidt relations across the optical disk of the M63 galaxy, and, in particular, close
to and beyond the isophotal radius, r25 . This exercise especially is of interest in the
bright UV region located in the outer disk of M63 at the galactocentric radius rgal =
1.36 r25 , whose stellar emission is solely dominated by the UV emission, so where
the star formation rate (SFR) metallicity, excitation, and gas density are potentially
lower. For this purpose, we use the photometric measurements obtained for the
FUV, 24 µm, and H I data. At each pointing defined for the CO radial mapping
and the CO mapping of the bright UV region, we compute the star formation rate
surface density, ΣSFR , the atomic gas surface density, ΣH I , and the molecular gas
surface density, ΣH2 .
The SFR surface density is determined using the calibration from Leroy et al.
(2008):
ΣSFR (M yr−1 kpc−2 ) = 8.1 × 10−2 FFUV (MJy sr−1 ) + 3.2 × 10−3 F24µm (MJy sr−1 )
(2.1)
which includes the FUV flux to measure the unobscured and recent star formation
through the emission of O and B stars, and the 24 µm flux that traces the obscured
FUV emission that is re-emitted in the FIR by dust grains.
41
The atomic gas surface density is measured with the calibration of Bigiel et al.
(2010):
ΣH I (M pc−2 ) = 0.020FH I (K km s−1 ) (2.2)
which includes a factor of 1.36 to reflect the presence of helium.

Finally, the molecular gas surface density (see Tables 2.1 and 2.2) is calculated
with our integrated CO(1–0) line fluxes using the calibration of Leroy et al. (2008):
ΣH2 (M pc−2 ) = 4.4 cos(i)FCO(1−0) (K km s−1 ) (2.3)
where the “standard” Galactic CO-to-H2 conversion factor, XCO = 2 × 1020 cm−2
(K km s−1 )−1 (Dickman et al. 1986), as well as the factor of 1.36 to account for
helium are adopted. The inclination correction by an angle i = 55◦ is taken into
account. This correction is also considered in the computation of the SFR and
atomic gas surface densities when deprojecting the FUV, 24 µm, and H I images
(Figure 2.7).
In Fig. 2.9, we plot the radial profiles of the star formation efficiency (SFE)
and the H2 -to-H I surface density ratio. The SFE is the star formation rate (SFR)
per unit of molecular gas, or the inverse of the gas depletion timescale, i.e. the time
required for present-day star formation to consume the gas reservoir. It is calculated
here as SFE = ΣSFR /ΣH2 . The various surface densities are computed using Eqs.
(2.1) to (2.3). The blue squares correspond to the pointings done to map CO along
the M63 major axis (the radial cut), and the red circles to those done to map CO
over the bright UV region in the outer disk of M63.
In the radial cut we consider only the pointings up to the isophotal radius, r25 ,
where we obtained convincing CO detections (17 pointings), and in the UV region
we retain pointings with 24 µm detections only (8 pointings) in order to have an
accurate estimation of the ΣSFR . We observe that the SFE, as measured over the
radial cut of M63, is roughly constant as a function of the galactocentric radius out
to the r25 limit. This is in line with the findings by Leroy et al. (2008) which we
now extend out to the isophotal radius. Beyond r25 , the two SFE detections tend to
show a drop of the SFE in the bright UV region, but upper limits in the other data
points prevent us from drawing definitive conclusions. The H2 -to-H I surface density
ratio shows a smooth decrease with the galactocentric radius, such that the inner
regions of the M63 disk are H2 -dominated, while the regions at rgal > 0.5 r25 end
up to be H I-dominated. This transition between a “mostly-H2 ” and a “mostly-H I”
interstellar medium (ISM) is found to be a well-defined function of local conditions
according to Leroy et al. (2008), occurring at characteristic galactocentric radius,
stellar and gas surface densities, hydrostatic gas pressure, and orbital timescale. At
the r25 limit, we observe a further and more severe drop of the H2 -to-HI surface
density ratio. Nevertheless, the behaviour of the radial profiles of the SFE and the
H2 -to-H I surface density ratio may well be inverted, if the trend toward an increasing
CO-to-H2 conversion factor with the galactocentric radius, as shown by Sandstrom
et al. 2013 (see their Fig. 7) for M63 over the radial cut out to rgal ∼ 0.7 r25 , is
confirmed.
42
rgal [kpc]
log SFE [yr−1] = log ΣSFR/ΣH2
0 5 10 15 20 25 30
−8.0
−8.5 Radial cut
−9.0 UV reg
−9.5
−10.0
−10.5
0 100 200 300 400 500 600
rgal [arcsec]
rgal [kpc]
0 5 10 15 20 25 30
1.5
1.0 Radial cut
log ΣH2 /ΣH I
0.5 UV reg
0.0
−0.5
−1.0
0 100 200 300 400 500 600

rgal [arcsec]
Figure 2.9: Radial profiles of the star formation efficiency (upper panel) and the H2 -to-H I
surface density ratio (bottom panel). The SFE is defined here as the star formation rate per unit
of molecular gas. The blue squares correspond to the 17 pointings done to map CO along the
M63 major axis (the radial cut) out to the isophotal radius, r25 (dashed line), and the red circles
correspond to the 8 pointings with 24 µm emission detection done to map CO over the bright UV
region at rgal = 1.36 r25 . The double data points plotted for the UV region at each rgal correspond
to the parallel pointings used to map this region (see Table 2.2).
43
10
8 leroy2008
ΣHI [M⊙pc−2]
6 this work
4
2
0
0 5 10 15 20 25
rgal [kpc]
103
leroy2008
102
ΣH2 [M⊙pc−2]
this work
101
100
10−1
0 5 10 15 20 25
rgal [kpc]
ΣSFR [M⊙yr−1kpc−2]
10−1 leroy2008
10−2 this work
10−3
10−4
10−5
0 5 10 15 20 25
rgal [kpc]
Figure 2.10: Comparison of ΣSFR , ΣHI and ΣH2 between this work and Leroy et al. (2008),
as a function of the galactocentric radius. Data points of this work correspond to each one of
the pointings used to maps the major axis of M63, and its UV region, whereas Leroy’s points
correspond to azimuthal averages over the entire disk.
Now, concerning the ΣSFR , ΣHI and ΣH2 , we can question how these surface den-
sities determined over single pointings along the radial major axis cut are represen-
tative of surface densities that Leroy et al. (2008) obtained by performing azimuthal
averages over the entire disk of M63. Such comparison is presented in Figure 2.10
The comparison of the respective molecular gas surface densities as a function of
the galactocentric radius is excellent. The SFR surface densities for the radial cut
show an offset toward higher values, but the respective ΣSFR radial profiles are the
same. Only the atomic gas surface density measurements seem to severely diverge
between the radial cut and the azimuthal averages. This difference comes from the
fact that the radial cut along the major axis crosses a spiral arm like structure,
which is very bright in H I and covers an H I emission peak in the outer parts of
the radial profile (see Figure 2.11). This certainly also explains the higher ΣSFR we
observe for the radial cut, where the star formation activity is slightly enhanced.
Finally, we compare the star formation rate surface densities with the atomic and
molecular gas surface densities, separately, in Kennicutt-Schmidt plots, in the two
top panels of Fig. 2.12. The blue squares correspond to the pointings done to map
CO along the M63 major axis, and the red circles to those done to map CO over the
44
Figure 2.11: HI map of M63 from Walter et al. (2008), with the observing CO pointings overlaid.
The radial cut along the major axis crosses a spiral arm like structure, very bright in HI.
bright UV region in the outer disk of M63 (similarly to Fig. 2.9). The two plots show
that the gas surface density of M63 in its radial cut is dominated by the molecular
gas, whereas the bright UV region is strongly dominated by the atomic gas, as also
observed in Fig. 2.9 (bottom panel). Interestingly, the H I surface density of the UV
region is very similar to the one observed along the radial cut, on the contrary to
its molecular gas surface density which is significantly lower. Finally, in the bottom
panel of Fig. 2.12 we show the star formation rate surface density as a function of
the total gas surface density, ΣH I+H2 , namely the sum of both ΣH I and ΣH2 .
To characterize the star formation in M63, we use the Kennicutt-Schmidt (K-S)
relation that relates the SFR surface density to the gas surface density. In all plots
of Fig. 2.12, we fit a power-law of the form ΣSFR = A(Σgas )N , which in a log-space
translates into a simple linear relation. The coefficient A traces the absolute star
formation efficiency (Kennicutt 1998b), and the slope N relates the star formation
rate to the gas density present and tells us how close are the two variables from
a linearity. We use the least-square bisector method, suitable for two independent
variables (Isobe et al. 1990), for the fit in the log-space and leave log A and N as
free parameters. It is important to note that for the fitting procedure only (not the
plotting), we re-scaled Σgas by a factor of 10, such that Σgas /10 M pc−2 . We made
this rescaling to minimize the covariance of log A and N in the fit4 , and to make
our results comparable to those of Bigiel et al. (2008) for M63. In their work they
used the same FUV and H I data as we did, but the CO data were taken from the
HERACLES CO(2–1) database (Leroy et al. 2009) with a 1100 resolution and a CO(2–
1)/CO(1–0) line ratio of 0.8. Our CO(1–0) data reach a much deeper sensitivity than
the HERACLES CO(2–1) data, which is essential to properly characterize the K-S
relation all over the optical disk and especially beyond the r25 limit.
4
This rescaling has no influence on the results of the slope N, only the coefficient A will need
to be corrected by the factor of 10.
45
0 0
−1 −1
−2 −2
log(ΣSFR) [M⊙ yr−1 kpc−2]

100% 100%
−3 −3
10% 10%
−4 −4
1% 1%
−5 −5
radial cut: log A=-1.23±0.05, N=2.85±0.15, σ=0.78 radial cut: log A=-1.93±0.01, N=1.06±0.01, σ=0.29
UV reg: log A=-2.13±0.22, N=4.42±0.52, σ=0.35 UV reg: log A=0.17±3.23, N=3.0±2.48 (indicative)
−6 −6
−1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5
log(ΣHI) [M⊙ pc−2] log(ΣH2 ) [M⊙ pc−2]
−1
−2
100%
−3
10%
−4
1%
−5
Radial cut, log A=-2.43±0.01, N=1.73±0.01, σ =0.31
UV region, log A=-2.2±0.23, N=4.7±0.6, σ =0.35
−1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5
log(ΣHI+H2 ) [M⊙ pc−2]
Figure 2.12: Kennicutt-Schmidt relations that relate the SFR surface density to the gas surface
density plotted for ΣH I (top-left panel), ΣH2 (top-right panel), and ΣH I+H2 (bottom panel). The
blue squares correspond to the 17 pointings done to map CO along the M63 major axis (the
radial cut) out to the isophotal radius, r25 , and the red circles correspond to the 8 pointings with
24 µm emission detection done to map CO over the bright UV region at rgal = 1.36 r25 . The K-S
relation is best parametrized by a power-law of the form ΣSFR = A(Σgas )N , which in a log-space
translates into a simple linear relation. The best-fitting bisector linear K-S relations are derived
separately for the radial cut (blue dashed-dotted lines) and for the UV region (red dashed lines).
The corresponding coefficients log A and slopes N can be found in the labels of each panel. The
black dotted lines represent “isochrones” of constant star formation efficiencies, indicating the level
of ΣSFR needed to consume 100%, 10%, and 1% of the total amount of gas within 108 years. The
vertical dashed line in the top-left panel corresponds to the ΣH I ∼ 9 M pc−2 threshold at which
the atomic gas saturates (Bigiel et al. 2008).
46
Table 2.6. Best-fitting bisector linear Kennicutt-Schmidt relations obtained for
M63 in log space.
This work Bigiel et al. 2008
Radial cut out to r25 UV region at 1.36 r25 Radial cut out to 0.68 r25
log A N σ log A N σ log A N σ
H2 -1.93 1.06 0.29 0.17† 3.0† ··· -2.22 0.92 0.10

HI+H2 -2.43 1.73 0.31 -2.20 4.7 0.35 -2.63 1.58 0.22
Note. — The Kennicutt-Schmidt relations are best parametrized by power-laws of the form
ΣSFR = A(Σgas )N , which in log space translates into linear relations. † Only indicative log A and N
values, since they rely on two data points only, which makes them very uncertain.
The best-fitting bisector linear relations, log ΣSFR = log A + N log Σgas , obtained
independently for the K-S relation all over the radial cut out to r25 and the UV
region in the outer disk of M63 are shown in Fig. 2.12. The errors we provide on log
A and N are 1 σ quotes of the nearly Gaussian distributions of log A and N obtained
through Monte-Carlo simulations of 5000 data sets created from the original data,
with random values generated within their error bars. The scatter (σ) in our fits
typically are of the order of 0.3 dex. The corresponding K-S relation results are
presented in Table 2.6 for both Σgas = ΣH2 and Σgas = ΣH I+H2 , and are compared
to the K-S relations found by Bigiel et al. (2008) over the radial cut of the M63 disk
out to rgal = 0.68 r25 only. The respective K-S relations obtained over the radial
cut have very comparable slopes N∼ 1.0 for ΣH2 and N∼ 1.65 for ΣH I+H2 , with our
slopes being systematically slighly steeper. They also have comparable coefficients
log A ∼ −2.0 for ΣH2 and log A ∼ −2.5 for ΣH I+H2 , with our coefficients being
systematically higher by about 0.2 − 0.3 dex, but still within the fit scatter. The
one-to-one linearity of the K-S relation, between the SFR surface density and the
molecular gas surface density is really solid, meaning that the gas is being consumed
at a nearly constant rate, as expected for ΣH2 .
On the other hand, when comparing these results to the K-S relations of the UV
region in the outer disk of M63, at the galactocentric radius rgal = 1.36 r25 , we note
that they are significantly different. The slope N = 4.7 of the K-S relation for ΣH I+H2
is considerably steeper in the external UV region. A similar trend is observed for
ΣH2 . It is true that it relies on two data points, but if we trust our 3 σ upper limits
on ΣH2 , we can hardly expect to reconcile these external UV region measurements
with the K-S relation observed over the radial cut, more especially as the derived
star formation rate surface densities are very reliable. It is very likely that the
CO-to-H2 conversion factor adopted over the external UV region is underestimated,
given the low metallicity expected in the outskirts of M63, as supported by the
metallicity gradient inferred in M63, and the observed trend for an increase in XCO
with decreasing metallicity (Sandstrom et al. 2013, Bolatto et al. 2013). While this
may induce a shift of the external UV region data points towards larger gas surface
densities, the slope of the K-S relation will remain unchanged, unless there is a
significant XCO gradient over the external UV region. As a result, a broken K-S
47
power-law between the inner and outer regions of the M63 disk seems to robustly
emerge. It suggests that the regime of star formation drastically changes beyond
the isophotal radius, r25 . Indeed, the slopes observed for the K-S relations in the
external UV region show a non-linear SFR regime, in clear contrast with the quasi-
linear SFR regime in the inner regions of the disk. What triggers the “quenching”
of the star formation in the outer regions of the M63 disk and, consequently, the
severe deviation from the commonly accepted SF relation as traced by the K-S
relation observed along the radial cut of M63? We can invoke the flaring of the outer
gas layers or, more speculative, the possible presence of high turbulence providing
support against gravitational collapse as suggested by Longmore et al. (2013).
To better appreciate the change in the star formation regime between the inner
and outer regions of the M63 disk, in Fig. 2.12 we also plot “isochrones” of constant
star formation efficiencies, indicating the levels of ΣSFR needed to consume 100%,
10%, and 1% of the total amount of gas (corrected by a factor of 1.36 to account
for helium) within 108 years. These isochrones can also be interpreted as constant
gas depletion timescales (i.e., the time needed to consume the total amount of gas)
of 108 , 109 , and 1010 years from top to bottom. We can see that when the total
gas is considered (H I + H2 ), for the radial cut of M63 the SFE settles the gas
consumption within 108 years to 1 to 10%, whereas the external UV region has a
much lower SFE with less than 1% of the gas being converted into stars within 108
years. The drop of the star formation efficiency beyond the isophotal radius, r25 ,
can also be appreciated in Fig. 2.9 (upper panel). The SFE could even be lower
in the external UV region, as the CO-to-H2 conversion factor may, in reality, be
up to 100 times larger (in the extreme case) than the “Galactic” value (Bolatto
et al. 2013), and so are the computed ΣH2 values, depending on the metallicity of
the external UV region which may well be as low as 10 − 20% of the local Galactic
ISM5 . Consequently, the star formation still occurs in the outer regions of the M63
disk, but at a very low efficiency, significantly lower than in the inner regions of the
disk.
2.1.4 Summary and conclusions
Deep CO(1–0) and CO(2–1) observations obtained on the IRAM 30 m telescope

of the M63 spiral galaxy characterized by a large XUV disk extending out to 2.5
times the optical equivalent radius are presented. We performed both a CO mapping
along the major axis of the M63 disk from the center out to the galactocentric radius
rgal = 57200 = 1.6 r25 and over a bright UV region in the outer disk of M63 at the
galactocentric radius rgal = 1.36 r25 . Our objective was to search for CO emission
and hence for molecular gas in the outer regions of the M63 disk beyond the optical
radius, where evidence of star formation is brought by both the XUV emission and
high H I column densities observed in these regions. We highlight the importance
of a CO detection in regions far away from the center of the galaxy, where the
metallicity, excitation, and gas density are supposed to be lower, and stars have
5
When extrapolating the metallicity gradient determined for M63, we get a metallicity of 20%
of the local Galactic ISM at rgal = 1.36 r25 , the galactocentric radius of the bright UV region.
48
more difficulty to form. To complement our CO observations, we used FUV, NUV,
Hα, 24 µm, and H I data from the literature. This allowed us to investigate the
Kennicutt-Schmidt relation across the galaxy and beyond the isophotal radius, r25 ,
in the bright UV region. Our main results are as follows:
The CO(1–0) emission is clearly detected along the major axis of the M63 disk
out to the isophotal radius, r25 , but not beyond. However, the CO(1–0) is again
detected in the bright UV region in the outer disk of M63 at rgal = 1.36 r25 . This
is the fourth molecular gas detection in the outskirts of nearby spiral galaxies. The
CO(2–1) emission is, on the other hand, confined to rgal = 0.68 r25 , and thus suggests
subthermal excitation in the outer regions of the M63 disk.
The radial profiles of the CO emission and other star formation tracers (Fig. 2.8)
show a severe drop with the galactocentric radius, on the contrary of the atomic
gas. Close to the r25 limit, CO and the star formation tracers begin to considerably
vanish and, beyond r25 , they all are practically absent with the exception of the faint
UV emission and H I. The UV region at rgal = 1.36 r25 , in which the CO emission
is detected, is characterized by FUV and NUV emission fluxes similar to the fluxes
observed at r25 , but stronger than the fluxes observed at rgal > r25 along the M63
major axis. This likely reflects a tight correlation between the CO and UV fluxes,
namely between the intensity of star formation and the amount of molecular gas,
and hence strongly suggests that the absence of CO detection at rgal > r25 , where
the XUV is weaker, is simply the result of the still too high CO detection threshold.
The external UV region is characterized by a very high H I flux with respect
to the measured CO flux. This leads us to speculate that H I is more likely the
precursor of H2 rather than the product of UV photo-dissociation, since it seems to
dominate in quantity. This is, however, true as long as substantial H2 is not hidden
in the outer disk regions of M63, which may be the case as we observe hints for an
excitation temperature decrease with the galactocentric radius, which may lead to
very weak CO lines.
With the integrated CO line flux measurements and the complementary data
from the literature, we derive SFR, H I, and H2 surface densities all along the major
axis of the M63 disk and in the external UV region. We observe that the gas surface
density along the radial cut is dominated by the molecular gas, whereas the UV
region is dominated by the atomic gas. The best-fit Kennicutt-Schmidt relations,
log ΣSFR = log A + N log Σgas , show a broken power-law from the inner to the outer
regions of the M63 disk (Fig. 2.12). Indeed, the almost linear K-S relation (with
a slope of nearly 1 in log space) observed over the radial cut, in the inner regions
of the disk characterized by high gas densities, cannot be extrapolated to the outer
disk regions. The later are characterized by a non-linear SFR regime (with a K-S
slope much higher than 1 in log space), perhaps owing to the flaring of the outer
gas layers. This is the first time that the K-S relation is quantified in the outskirts
of a spiral galaxy, i.e, in low gas density environments. At a molecular gas surface
density as low as ΣH2 = 0.35 M pc−2 , well below all the determined H2 surface
densities referenced in spiral galaxies so far (Bigiel et al. 2008, 2011), star formation
still occurs spontaneously.
49
The change in the star formation regime between the inner and outer regions of
the M63 disk can also be appreciated by the difference in star formation efficiencies.
Indeed, along the major axis of the disk out to the isophotal radius the SFE settles
the gas consumption within 108 years to 1% to 10%, whereas in the UV region much
less than 1% of the gas is converted into stars within 108 years. Consequently, star
formation still occurs in the outer regions of the disk, but at a very low efficiency.
2.1.5 Publication
This entire work done on M63 was submitted to A& A in December 2013, and
accepted for its publication in April 2014. The final publication is presented here.
50
A&A 566, A147 (2014) Astronomy
DOI: 10.1051/0004-6361/201323330 &

c ESO 2014 Astrophysics
CO map and steep Kennicutt-Schmidt relation in the extended

UV disk of M 63
M. Dessauges-Zavadsky1 , C. Verdugo2 , F. Combes2 , and D. Pfenniger1
1
Observatoire de Genève, Université de Genève, 51, Ch. des Maillettes, 1290 Sauverny, Switzerland
e-mail: [miroslava.dessauges;daniel.pfenniger]@unige.ch
2
Observatoire de Paris, LERMA, 61 Av. de l’Observatoire, 75014 Paris, France
e-mail: [celia.verdugo;francoise.combes]@obspm.fr
Received 23 December 2013 / Accepted 5 April 2014
ABSTRACT
Results from the UV satellite GALEX revealed surprisingly large extensions of disks in some nearby spiral galaxies. While the
Hα emission, the usual tracer of star formation, drops down at the border of the isophotal radius, r25 , the UV emission extends out to
3 to 4 times this radius and often covers a significant fraction of the H I area. M 63 is a remarkable example of a spiral galaxy with one
of the most extended UV disks, so it offers the opportunity to search for the molecular gas and characterize the star formation in outer
disk regions as revealed by the UV emission. We obtained deep CO(1–0) and CO(2–1) observations on the IRAM 30 m telescope
along the major axis of the M 63 disk from the center out to the galactocentric radius rgal = 1.6 r25 and over a bright UV region at
rgal = 1.36 r25 . CO(1–0) is detected all along the M 63 major axis out to r25 , and CO(2–1) is confined to rgal = 0.68 r25 , which may
betray lower excitation temperatures in the outer disk. CO(1–0) is also detected in the external bright UV region of M 63. This is the
fourth molecular gas detection in the outskirts of nearby spirals. The radial profiles of the CO emission and of the Hα, 24 μm, NUV
and FUV star formation tracers and H I taken from the literature show a severe drop with the galactocentric radius, such that beyond
r25 they are all absent with the exception of a faint UV emission and H I. The CO emission detection in the external UV region, where
the UV flux is higher than the UV flux observed beyond r25 , highlights a tight correlation between the CO and UV fluxes, namely the
amount of molecular gas and the intensity of star formation. This external UV region is dominated by the atomic gas, suggesting that
H I is more likely the precursor of H2 rather than the product of UV photodissociation. A broken power law needs to be invoked to
describe the Kennicutt-Schmidt (K-S) relation of M 63 from the center of the galaxy out to rgal = 1.36 r25 . While all along the major
axis out to r25 , the K-S relation is almost linear (with a slope of nearly 1 in log space), in the external UV region the SFR regime is
highly nonlinear and characterized by a steep K-S relation (with a slope much higher than 1 in log space) and very low star formation
efficiency.
Key words. galaxies: star formation – ultraviolet: galaxies – galaxies: ISM – submillimeter: ISM – galaxies: evolution
1. Introduction galactocentric radii. Recent star formation within such environ-

ments was detected in Hα (the principal star formation tracer
The study of star formation in the outer regions of disks of nor- over the years) and broad-band observations for a few galax-
mal spiral galaxies has gained interest in the past few years, ies: NGC 628, NGC 1058, NGC 6946 (Ferguson et al. 1998),
mainly because they are low-metallicity environments (Henry M 31 (Cuillandre et al. 2001), and NGC 6822 (de Blok &
& Worthey 1999), resembling the conditions of early stages of Walter 2003). However, the GALEX far-UV (FUV) and near-
spiral galaxies and high-redshift galaxies. These regions are also UV (NUV) data demonstrate that Hα observations still fail to
known to have low star formation rates (Dong et al. 2008; Bigiel detect a significant population of moderate-age stars in the out-
et al. 2010; Alberts et al. 2011). Moreover, there is growing ev- ermost disks of spiral galaxies, since Hα traces more recent star
idence of cold gas accretion in the local Universe, both through formation episodes. Indeed, UV-bright disks extending up to 3
the arrival and merging of gas-rich satellites and through gas in- to 4 times the optical radius have been reported in about 30%
fall from the intergalactic medium. This new gas could be de- of spiral galaxies, with the most remarkable examples: M 63,
posited in the outer regions of galaxies and form reservoirs for M 83, NGC 2841, and NGC 4625 (Thilker et al. 2005; Gil de Paz
replenishing the inner parts and feeding star formation (Sancisi et al. 2005, 2007). These extended UV emission (XUV) disks
et al. 2008), making outer regions good laboratories for scanning cover a significant fraction of the area detected at 21 cm wave-
the interface between galaxies and the surrounding intergalactic length, with some correspondence between the position of the
gas. brightest UV complexes and peaks in the atomic gas distribu-
With its 1.25 degree field of view and sensitivity to stellar tion. The measured FUV − NUV colors are generally consistent
populations younger than a few hundred Myr, the UV satel- with young populations of O, and predominantly B stars, char-
lite GALEX (Galaxy Evolution Explorer) is well suited to ad- acterized by ages from a few Myr up to 400 Myr.
dress the question of star formation in spiral galaxies at large
The confirmed occurrence of recent and ongoing star for-

Based on observations carried out with the IRAM 30 m telescope. mation in the outer disks of normal spiral galaxies has several
IRAM is supported by INSU/CNRS (France), MPG (Germany), and important implications. First, it supports the presence of molec-
IGN (Spain). ular gas in the outskirts of spirals, since stars are formed within
Article published by EDP Sciences A147, page 1 of 12
A&A 566, A147 (2014)
molecular clouds. Second, this suggests the presence of large

reservoirs of hydrogen in the form of H2 , which may contribute 42◦ 10
to the baryonic dark matter of spiral galaxies. Third, it offers
the ideal place to study the unresolved issue of the atomic hy-
drogen gas origin: is H I mainly a product of the star formation
process; i.e., does it result from the photodissociation of H2 by
42◦ 05
the UV flux radiation emanated from newly formed stars (Allen
et al. 1986, 2004; Smith et al. 2000), rather than mainly being
DEC. (J2000)
a precursor to it? Fourth, the presence of recently formed stellar
complexes at large galactocentric radii also provides a simplified
laboratory for determining the star formation threshold, namely 42◦ 00
the minimum gas surface density required for star formation
to occur spontaneously (Kennicutt 1989; Martin & Kennicutt
2001). Fifth, it allows investigating the star formation in qui-
escent and low-metallicity environments that may affect the star
formation density and the initial mass function. 41◦ 55
In this paper we aim to detect the molecular gas expected
in the outskirts of spiral galaxies because of the star formation
discovered from the XUV observations. The M 63 spiral galaxy, 30s 13h 16m 00s 30s 13h 15m 00s
known to have an XUV disk, is selected for this work. The search R.A. (J2000)
for molecular gas in galaxies is difficult, since we cannot detect
Fig. 1. False three-color (red/green/blue) composite image of M 63 from
cold H2 directly. Instead, the second most common molecule, Gil de Paz et al. (2007). The image was produced using the arcsinh
CO, is used as a proxy. Molecular gas has been detected in many function, which allows showing faint structures while simultaneously
galaxies from the mapping of the CO emission. It is now well preserving brighter structures in the field, such as the spiral arms of
established that the CO emission is the strongest in the central large galaxies (Lupton et al. 2004). The green ellipse represents the
regions of spiral galaxies (e.g., Young & Scoville 1991), but then B-band 25 mag arcsec−2 isophote limit, r25 . The red solid line shows
falls off, as does the blue stellar light, with the galactocentric the CO mapping we performed along the major axis of M 63 from the
radius (e.g., Young & Scoville 1982; Young et al. 1995). The center out to rgal = 572 = 1.6 r25 , where 27 single pointings were
questions “is this CO emission drop real, or does it reflect the aligned with a spatial sampling of 22 . The red box encloses the bright
difficulty of detecting the molecular gas in the outer regions of UV region at rgal = 483 = 1.36 r25 which we mapped with 6 × 2 =
12 pointings with the same spatial sampling. The CO(1–0) emission is
spiral galaxies” remain open. For example, CO emission lines
observed out to the isophotal radius along the radial cut and in the bright
are known to be weaker in media characterized by low metal- UV region, while the CO(2–1) emission is confined to 0.68 r25 (dashed
licities, low gas temperatures, low excitations, and low gas den- green ellipse).
sities, even if substantial molecular gas is present (e.g., Allen
1996; Combes & Pineau Des Forets 2000), and thus in media
that may well be potentially representative of outer disk regions. 2. M 63 characteristics
Schruba et al. (2011) used a very effective stacking technique Classified as SA(rs)bc and located at 10.1 Mpc (Leroy et al.
on HERACLES CO(2–1) data of nearby spiral galaxies, where 2009), M 63 (or NGC 5055) looks like a typical spiral galaxy,
they stack CO spectra across many sightlines by assuming that representative of a large class of local spirals, with no imme-
the mean H I and CO velocities are similar, to highlight the pres- diate neighbour, which excludes the potentiality of a galaxy in
ence of the faintest CO emission. They focused their analysis on interaction. However, M 63 is not very ordinary, since it is a re-
data stacked in bins of galactocentric radius and found that the markable example of a nearby spiral with a bright and XUV disk.
CO radial profile follows a remarkably uniform exponential de- In Fig. 1 we show the GALEX NUV and FUV color-composite
cline with a scale length of ∼0.2 r25 . But even though this has image of M 63, where an extensive population of UV-bright star-
been shown to be a very effective technique, it still fails to detect forming regions and stellar clusters (tracing the O and B stars)
CO in the farthest out regions of galactic disks, where they only is revealed. The M 63 NUV and FUV surface-brightness pro-
reach 3σ upper limits. Specific very deep CO observations are files show a smooth decrease in the UV emission out to 700 in
thus required to trace the molecular gas in the outer disk regions the equivalent radius (defined as the square root of the product
of spiral galaxies, which is precisely the objective of this work. of the half-minor axis times the half-major axis), namely out to
In Sect. 2 we present the characteristics of the M 63 spi- 2.5 times the optical equivalent radius (Gil de Paz et al. 2007).
ral galaxy, selected for deep CO observations in the outer disk The major axis radius of the B-band 25 mag arcsec−2 isophote,
regions. We qualify the advantages of M 63 in the context of the so-called isophotal radius, is equal to r25 = 354 = 17.4 kpc
XUV galaxies and cite the previous works done on M 63. In in M 63. About 30% of local spiral galaxies have XUV disks as
Sect. 3 we describe the CO observations performed on the IRAM compared to their optical r25 disks, among which M 63 is one
30 m telescope and show the corresponding results in Sect. 4. In of the most extreme cases (Thilker et al. 2005; Gil de Paz et al.
−1
Sect. 5 we discuss the radial profiles obtained for the acquired 2005, 2007). A metallicity gradient of −0.59 dex r25 was re-
CO measurements in comparison to other star formation tracers ported by Moustakas et al. (2010) for the calibration of Pilyugin
(FUV, NUV, Hα, 24 μm) and H I, and investigate the Kennicutt- & Thuan (2005), which gives 12 + log(O/H) = 8.59 ± 0.07 for
Schmidt relations across the galaxy and beyond the optical ra- the central metallicity of M 63.
dius. Summary and conclusions are given in Sect. 6. Throughout The 21 cm observations of M 63 also show the presence of a
the paper we adopt the “standard” Galactic CO(1–0)-to-H2 con- very large, warped gaseous disk extending out to 40 kpc in the
version factor XCO = 2 × 1020 cm−2 (K km s−1 )−1 (Dickman et al. major axis radius (e.g., Battaglia et al. 2006). The warp starts
1986) and include a correction for helium. around r25 and is exceptionally extended and symmetric. The
A147, page 2 of 12
M. Dessauges-Zavadsky et al.: CO and Kennicutt-Schmidt relation in the XUV disk of M 63
measured H I column densities are higher than 1020 cm−2 up to mode with the maximum symmetrical azimuthal wobbler throw
70% of the XUV disk. These high H I column densities plus the of 240 allowed, corresponding to 11.8 kpc in projected distance
extended UV emission, both suggest the presence of molecu- for this spiral galaxy located at a distance of 10.1 Mpc. The total
lar gas out to large galactocentric radii. Indeed, recent studies on-source exposure time obtained per pointing in the M 63 disk
have demonstrated that strong H I emission is on average a good map and the UV region map is listed in Tables 1 and 2 (Col. 3),
tracer of regions rich in molecular gas (e.g., Crosthwaite et al. respectively. Only scans with a system temperature lower than
2002; Nieten et al. 2006), and the extended UV emission unde- 400 K at 3 mm were retained for analysis.
niably betrays the presence of relatively young stars, hence of The data were reduced with the CLASS software from the
the molecular gas necessary for their formation. GILDAS package. All the spectra obtained with the two re-
Molecular gas was looked for in M 63 in the context of ceivers tuned on the 12 CO(1–0) line and corresponding to scans
the BIMA/SONG survey, the first systematic imaging survey of at the same pointings were summed up without any smoothing.
CO(1–0) emission from the centers and disks of nearby galax- On the other hand, the spectra obtained with the two receivers
ies (Helfer et al. 2003, see their Fig. 43). The reported high- tuned on the 12 CO(2–1) line were first Hanning-smoothed to a
resolution CO measurements are confined to the very central resolution of 2.6 km s−1 , since the expected average CO line full
area of the M 63 optical disk, rgal = 96 = 0.3 r25 along the width half maximum is approximately 10 km s−1 , before being
major axis, because of the lack of sensitivity of these data at the summed up when corresponding to scans at the same pointings.
border of the map with a primary beam gain drop by a factor of 2. No baseline subtraction was performed on individual spectra be-
Single-dish spectra from the FCRAO Extragalactic CO Survey fore summing because we simply did a linear sum. In Tables 1
(Young et al. 1995, see their Fig. 87), despite their moderate sen- and 2 (Cols. 4 and 7), we list the achieved 1σ rms in mK at
sitivity and low resolution, show that the CO emission extends 3 mm and 1 mm for each pointing of the M 63 disk map and the
over a larger area than mapped by BIMA/SONG, reaching a two UV region map, respectively1. These rms values were obtained
times larger galactocentric radius, but still not extending up to with a baseline subtraction of degree 0 and with windows set
the optical disk limit r25 . The signal measured at rgal = 180 is to 300−400 km s−1 , both at 3 mm and 1 mm, as defined by the
equal to ∼2 K km s−1 , which leaves room for a CO line flux velocity positions of detected CO(1–0) and CO(2–1) lines2 . At
decrease by a factor of 10 to 20 in the outermost XUV disk 3 mm we obtained on average rms values between 4 mK and
(depending on the CO line width), when aiming at signals as 21 mK in the M 63 disk map, while for seven pointings around
weak as 10 mK, or weaker, in these extreme regions. More re- the isophotal radius, r25 , and for all the pointings of the UV re-
cently, Leroy et al. (2009) mapped the CO(2–1) line in M 63 gion map, we pushed the rms limit down to 4 mK to 7 mK.
over the full optical disk, as part as the HERACLES survey on
the IRAM 30 m telescope, and obtained reliable measurements
out to rgal = 0.68 r25 . 4. Results
As summarized in Sect. 2, the CO(1–0) and CO(2–1) lines were
3. Observations and data reduction previously mapped over the full optical disk of M 63 in the con-
text of various CO surveys of nearby galaxies. However, none of
The CO observations were performed with the IRAM 30 m
these observations reaches the sensitivity of our data, in particu-
millimeter-wave telescope at Pico Veleta, Spain, during a first
lar around the isophotal radius, r25 , and beyond. High sensitivity
run on September 10–17, 2007 under poor weather conditions,
can be achieved by mapping CO through individual beam point-
and during a second run on November 2, 16, and 27 and
ings, which is a technique that proves to be very efficient for
December 1, 2007 under good-to-excellent weather conditions.
obtaining deep observations of specific, not extended, areas of a
We used four single-pixel heterodyne receivers, simultaneously,
galaxy.
two centered on the 12 CO(1–0) line at 115.271 GHz, and two on
The results of our CO emission mapping of M 63 along
the 12 CO(2–1) line at 230.538 GHz. The telescope half-power
the major axis of its disk and over the bright UV region at
beam widths at these two frequencies are 22 and 11 , respec-
rgal = 1.36 r25 are presented in Tables 1 and 2, respectively. We
tively. The data were recorded using the VESPA autocorrelator
provide the CO(1–0) and CO(2–1) line full widths half maxi-
with 640 MHz bandwidth and 1.25 MHz resolution at 3 mm,
mum, FWHM (Cols. 5 and 8) and the integrated CO(1–0) and
and two 1 MHz filter banks (512 channels each) at 1 mm. The
CO(2–1) line fluxes (Cols. 6 and 9) at each pointing of the
resulting velocity coverage at 115.271 GHz is 1665 km s−1 with
mapping, as determined from fitting Gaussian functions to the
a spectral resolution of 3.2 km s−1 . The corresponding values at
CO profiles obtained by summing up all spectra corresponding
230.538 GHz are 666 km s−1 and 1.3 km s−1 .
to all scans per pointing. The CO(1–0) and CO(2–1) line FWHM
First, we performed a CO mapping along the major axis of
are in the range from 9 km s−1 to 53 km s−1 and from 6 km s−1
the M 63 disk from the center of the galaxy located at (J2000)
to 39 km s−1 , respectively. They are greater than those expected
RA = 13h15m49.3s, Dec = +42◦ 01m45.4s out to the galacto-
for individual giant molecular clouds (GMCs) that have typical
centric radius rgal = 572 by aligning 27 single pointings with
line widths of 10.4 km s−1 (Solomon et al. 1987), except for
a spatial sampling of 22 in the radial direction sustaining a po-
a few pointings around r25 and over the UV region. This im-
sition angle PA = 105◦ . Second, we mapped a bright UV re-
plies that mostly an ensemble of molecular clouds is emitting
gion in the outer regions of the M 63 disk centered on (J2000)
per beam of 1 kpc and 0.5 kpc at 3 mm and 1 mm, respectively.
RA = 13h15m07.0s, Dec = +42◦ 00m00.0s and located at the
galactocentric radius rgal = 483 with 6 × 2 = 12 pointings fol- 1
The efficiencies of the ABCD receivers on the IRAM 30 m tele-
lowing the sequence from (−66 ; 0 ) to (+44 ; +22 ) offsets scope are 5.9 Jy K−1 at 3 mm and 7.2 Jy K−1 at 1 mm (see the on-
relative to the central coordinates with a 22 spatial sampling line IRAM wiki pages at http://www.iram.es/IRAMES/mainWiki/
in the right-ascension direction and the same in the declina- Iram30mEfficiencies).
tion direction. In Fig. 1 we show the false-color GALEX image 2
Except for the very few inner pointings at rgal = 0, 22 , and 44 , for
of M 63 on top of which the regions targeted for CO emission which a larger window was set, given the larger widths of the CO(1–0)
are plotted. Observations were performed in wobbler-switching and CO(2–1) lines.
A147, page 3 of 12
A&A 566, A147 (2014)
Table 1. CO mapping along the major axis of the M 63 disk.
CO(1–0) CO(2–1)
rgal rgal Time rmsa FWHMb Fc rmsa FWHMb Fc r2,1 d MH2 e ΣH2 f
( ) (r25 ) (min) (mK) −1
(km s ) (K km s−1 ) (mK) −1
(km s ) (K km s−1 ) (106 M ) (M pc2 )
0 0 31 11 200 49.8 ± 2.2 16 192 39.8 ± 3.1 0.24 321.8 ± 14.2 125.7 ± 5.6
22 0.06 8 18 53 24.5 ± 0.9 28 27 12.6 ± 0.8 0.16 158.3 ± 5.8 61.8 ± 2.3
44 0.12 8 19 41 16.4 ± 0.8 33 21 9.0 ± 0.7 0.17 106.0 ± 5.2 41.4 ± 2.0
66 0.19 8 19 23 9.8 ± 0.4 32 16 4.2 ± 0.5 0.13 63.3 ± 2.6 24.7 ± 1.0
88 0.25 8 18 20 6.3 ± 0.4 32 20 3.4 ± 0.6 0.17 40.7 ± 2.6 15.9 ± 1.0
110 0.31 8 19 26 12.6 ± 0.5 30 21 8.2 ± 0.6 0.20 81.4 ± 3.2 31.8 ± 1.3
132 0.37 8 21 22 5.1 ± 0.5 35 20 2.6 ± 0.7 0.16 33.0 ± 3.2 12.9 ± 1.3
154 0.44 23 13 29 5.3 ± 0.4 22 39 2.5 ± 0.9‡ 0.15 34.3 ± 2.6 13.4 ± 1.0
176 0.50 23 13 19 5.4 ± 0.2 24 27 2.8 ± 0.6 0.16 34.9 ± 1.3 13.6 ± 0.5
198 0.56 23 13 21 1.8 ± 0.3 22 18 0.82 ± 0.40‡ 0.14 11.6 ± 1.9 4.5 ± 0.8
220 0.62 31 11 16 0.98 ± 0.17 11† 10 <0.33 <0.11 6.3 ± 1.1 2.5 ± 0.4
242 0.68 101 6 11 0.30 ± 0.06 8 6 0.17 ± 0.05 0.17 1.9 ± 0.4 0.76 ± 0.15
264 0.75 101 6 14 0.45 ± 0.08 4† 10 <0.12 <0.08 2.9 ± 0.5 1.1 ± 0.2
286 0.81 55 10 21 0.88 ± 0.21 9† 10 <0.27 <0.10 5.7 ± 1.4 2.2 ± 0.5
308 0.87 47 11 14 0.81 ± 0.15 11† 10 <0.33 <0.13 5.2 ± 1.0 2.0 ± 0.4
330 0.93 63 7 22 0.85 ± 0.15 9 34 0.38 ± 0.31‡ 0.14 5.5 ± 1.0 2.1 ± 0.4
352 0.99 155 5 9 0.23 ± 0.04 6 34 0.29 ± 0.21‡ 0.38 1.5 ± 0.3 0.58 ± 0.10
374 1.06 109 4† 10 <0.12 4† 10 <0.12
396 1.12 117 4† 10 <0.12 3† 10 <0.09
418 1.18 93 5† 10 <0.15 5† 10 <0.15
... ...
473 1.34 381 2† 10 <0.06 2† 10 <0.06 <0.39 <0.15
Notes. The radius of the B-band 25 mag arcsec−2 isophote, the so-called isophotal radius, is equal to r25 = 354 = 17.4 kpc in M 63. It is often
used as a reference to express the relative galactocentric radius, rgal . The last separate line of the Table gives the values obtained when summing
all scans at the 10 outermost pointings from 1.06 r25 to 1.6 r25 with no CO detection and by smoothing the resulting 3 mm and 1 mm spectra to a
resolution of 9.7 km s−1 and 10.4 km s−1 , respectively. (a) Rms noises at, respectively, 3 mm and 1 mm, measured per channel of 3.2 km s−1 and
1.3 km s−1 . The values marked with † correspond to rms noises measured per smoothed channel of 9.7 km s−1 and 10.4 km s−1 , respectively. (b) Full
widths at half maximum of, respectively, CO(1–0) and CO(2–1) lines as determined from fitting Gaussian profiles. (c) Integrated CO(1–0) and
CO(2–1) line fluxes as determined from fitting Gaussian profiles. The values marked with ‡ correspond to potentially less reliable measurements,
because of a CO line detection at 2σ only. Upper limits are 3σ and are calculated assuming a 10 km s−1 line width. (d) CO luminosity ratios
defined as r2,1 = LCO(2−1) /LCO(1−0) , needed to correct the lower Rayleigh-Jeans brightness temperature of the 2–1 transition relative to 1–0. We
would like, however, to stress that CO(1–0) and CO(2–1) do not exactly map the same regions, because the CO(1–0) beam area equals 4×
the CO(2–1) beam area. The CO luminosity is calculated using the formula (3) from Solomon et al. (1997). (e) H2 masses calculated from the
CO(1–0) luminosity and by adopting the “standard” Galactic CO-to-H2 conversion factor, XCO = 2 × 1020 cm−2 (K km s−1 )−1 (Dickman et al.
1986). The applied formula is: MH2 (M ) = 4.4LCO(1−0) (K km s−1 pc2 ), where a factor of 1.36 is included to account for helium. ( f ) H2 surface
densities calculated from the integrated CO(1–0) line flux and by adopting the CO-to-H2 conversion factor given under. (e) The applied formula is:
ΣH2 (M pc−2 ) = 4.4 cos(i)FCO(1−0) (K km s−1 ), where i = 55◦ is the inclination of M 63 (Leroy et al. 2008).
Table 2. CO mapping over the bright UV region at 1.36 r25 .
CO(1–0) CO(2–1)
Offsets Time rmsa FWHMb Fc rmsa FWHMb Fc r2,1 d MH2 e ΣH2 f
( ) ( ) (min) (mK) (km s−1 ) (K km s−1 ) (mK) (km s−1 ) (K km s−1 ) (106 M ) (M pc2 )
−66 0 118 5 10 <0.15 6 10 <0.18 <0.97 <0.38
−66 +22 117 5 11 0.14 ± 0.06‡ 6 10 <0.18 <0.39 0.90 ± 0.39 0.35 ± 0.15
−44 0 186 4 10 <0.12 4 10 <0.12 <0.78 0.30
−44 +22 164 4 10 <0.12 6 10 <0.18 <0.78 0.30
−22 0 171 4 14 0.21 ± 0.06 5 10 <0.15 <0.22 1.36 ± 0.39 0.53 ± 0.15
−22 +22 148 4 10 <0.12 6 10 <0.18 <0.78 <0.30
0 0 164 4 10 <0.12 5 10 <0.15 <0.78 <0.30
0 +22 148 5 10 <0.15 6 10 <0.18 <0.97 <0.38
+22 0 123 5 10 <0.15 7 10 <0.21 <0.97 <0.38
+22 +22 116 5 10 <0.15 8 10 <0.24 <0.97 <0.38
+44 0 63 7 10 <0.21 9 10 <0.27 <1.36 <0.53
+44 +22 62 7 10 <0.21 8 10 <0.24 <1.36 <0.53
All offsets 1932 1 23 0.11 ± 0.02 2 10 <0.06 <0.17 0.71 ± 0.13 0.28 ± 0.05
Notes. The last separate line gives the values obtained when summing all scans at the 12 offset pointings. (a) Rms noises at, respectively, 3 mm
and 1 mm, measured per channel of 3.2 km s−1 and 1.3 km s−1 . (b) Full widths at half maximum of, respectively, CO(1–0) and CO(2–1) lines as
determined from fitting Gaussian profiles. (c) Integrated CO(1–0) and CO(2–1) line fluxes as determined from fitting Gaussian profiles. The value
marked with ‡ corresponds to a potentially less reliable measurement, because of the CO line detection at barely 3σ. Upper limits are 3σ and are
calculated assuming a 10 km s−1 line width. (d) +(e) +( f ) CO luminosity ratios, H2 masses, and H2 surface densities, respectively, calculated using
the same prescriptions as in Table 1.
A147, page 4 of 12
Fig. 2. Mosaic of CO spectra obtained when mapping the M 63 disk along its major axis from the center of the galaxy out to the isophotal
radius, r25 . The CO(2–1) spectra (red thin line) are overplotted on the CO(1–0) spectra (black thick line). While CO(1–0) is clearly detected up
to the optical radius, CO(2–1) appears to be confined to rgal < ∼ 0.68 r25 . The bottom right hand panel shows the CO(1–0) and CO(2–1) spectra
obtained when summing all scans at the ten outermost pointings from rgal = 1.06 r25 to 1.6 r25 and smoothed to a resolution of 9.7 km s−1 and
10.4 km s−1 , respectively. No CO emission is detected at these outermost pointings.
In Tables 1 and 2 we also list the molecular gas masses per point- all scans per pointing of the M 63 disk map along the major axis,
ing (Col. 11), as derived using the “standard” Galactic CO(1−0)- while Fig. 3 shows the resulting spectrum obtained by summing
to-H2 conversion factor XCO = 2 × 1020 cm−2 (K km s−1 )−1 up all spectra of all scans at the 12 pointings used to map the
(Dickman et al. 1986) and including a correction factor of external UV region.
1.36 for helium. The inferred masses are in the range MH2 = The radial mapping (Fig. 2) clearly shows a detection of the
(1.5−322) × 106 M . Figure 2 shows the summed-up spectra of CO(1–0) emission out to the galactocentric radius rgal = 352 ,
A147, page 5 of 12
A&A 566, A147 (2014)
Table 3. Public ancillary data of M 63.
Band or line Telescope FWHM Reference

( )
CO(2–1) IRAM 13.4 Leroy et al. (2009)
FUV GALEX 4.3 Gil de Paz et al. (2007)
NUV GALEX 5.3 Gil de Paz et al. (2007)
Hα KPNO 0.38 Kennicutt et al. (2008)
24 μm Spitzer 6 Dale et al. (2009)
H I (21 cm) VLA 6 Walter et al. (2008)
FCO(2−1) < 0.06 K km s−1 , obtained when summing up all the

spectra corresponding to all scans at the 12 pointings.
Fig. 3. CO(1–0) spectrum obtained by summing up all spectra at the
12 pointings used to map the bright UV region at rgal = 1.36 r25 . There
is a clear CO emission detection in this region of M 63 well beyond the 5. Discussion
optical radius. CO(2–1) remains undetected.
A few hundred nearby galaxies have been the object of intensive
surveys for molecular gas (e.g., Young & Scoville 1991; Young
et al. 1995; Helfer et al. 2003; Leroy et al. 2009). However, for
only a very few have deep CO searches in their outer disk regions
namely out to the isophotal radius r25 = 354 = 17.4 kpc. A se- been undertaken. Here, we show the existence of molecular gas
vere drop in the CO flux is nevertheless observed as a function of up to the isophotal radius, r25 , and beyond in the nearby spiral
the galactocentric radius. Beyond r25 and out to 1.6 r25 , the limit galaxy M 63. This is the fourth such a detection of molecular
of our CO search along the M 63 major axis, no CO emission is gas in the outskirts after the spiral galaxies NGC 4414 (Braine &
detected anymore (Table 1). Even by summing up all the spectra Herpin 2004), NGC 6946 (Braine et al. 2007), and M 33 (Braine
corresponding to all scans at the ten outermost pointings and by et al. 2010). To analyze the impact of our observations in the
smoothing the resulting spectrum to a resolution of 9.7 km s−1 , context of star formation in the outer disk of M 63, we present
i.e., similar to the typical FWHM of CO(1–0) lines detected at the radial profiles of our CO observations along with comple-
the outermost pointings, no CO emission is observed. We do, mentary data of other star formation tracers (FUV, NUV, Hα,
however, derive a stringent 3σ upper limit on the integrated 24 μm) and H I, and we investigate the behavior of the Kennicutt-
CO(1–0) line flux at rgal > r25 of FCO(1−0) < 0.06 K km s−1 , Schmidt relations across the galaxy and beyond r25 in the bright
when assuming a line width of 10 km s−1 . This is equivalent to UV region. Characterizing the Kennicutt-Schmidt relation in the
a molecular gas mass of MH2 < 3.9 × 105 M . On the other outer disk regions, namely in environments with low metallici-
hand, the CO(2–1) emission is securely detected only out to ties, gas temperatures, excitations, and gas densities, which are
rgal = 242 , i.e., 0.68 r25, in agreement with the CO(2–1) map- all properties hostile to star formation, is our main interest.
ping by Leroy et al. (2009), while there are two very tentative
detections at 2σ at the pointings rgal = 0.93 r25 and r25 (Table 1). 5.1. Radial profiles
The CO(2–1) line thus appears to be excited over about 2/3
of the optical disk. The corresponding CO luminosity ratios, To compare the CO distribution in M 63 along the major axis of
r2,1 = LCO(2−1) /LCO(1−0) , can be found in Table 1 (Col. 10). They its disk to H I and other star formation tracers’ distributions, we
vary randomly between 0.13 and 0.24 (in the center of M 63) built comparative radial profiles of the CO, H I, FUV, NUV, Hα,
along the galactocentric radius, without showing any radial evo- and 24 μm emission. Altogether they provide a complete view
lution. The absence of any clear detection of the CO(2–1) line of the past and ongoing star formation that took or takes place
beyond rgal = 0.68 r25 suggests that the subthermal excitation in the galaxy and of the ingredients needed to sustain the star
sets in around this galactocentric radius. However, this has to be formation: molecular and atomic gas. The respective data are
considered with caution because the major axis pointings probe taken from the literature and are described in Table 3. In Fig. 4
only a small portion of the M 63 disk at each radius and the Leroy we show the multiwavelength views of the bright UV region at
et al. (2009) CO(2–1) map of the entire M 63 disk is not very rgal = 1.36 r25 with, from left to right, the NVU, 24 μm, and H I
deep and because there seems still to be plenty of star-forming 21 cm emission.
regions at rgal > 0.68 r25 with high enough densities (locally) to As these data were obtained with different instruments and
thermalize the CO(2–1) line (see Fig. 1). hence at various FWHM resolutions, we first convolved all these
Whereas our survey for CO emission along the major axis data to the resolution of our CO(1–0) data which is equal to the
of the M 63 disk seems to indicate that the molecular gas is half-power beam width of the IRAM 30 m telescope at 3 mm,
confined to the isophotal radius r25 of the galaxy, we find a i.e., 22 . The images were then rotated by 20◦ in a north-to-
convincing CO(1–0) detection in the selected bright UV re- west direction to have the major axis in the horizontal axis,
gion at rgal = 1.36 r25 . The sum of all spectra corresponding and deprojected by correcting for a 55◦ inclination angle to
to all scans at the 12 pointings used to map the UV region bring the galaxy face-on. Finally, we did aperture photometry
leads to a CO(1–0) detection at 5.5σ with an integrated line flux on these processed images with apertures of 22 diameter us-
FCO(1−0) = 0.11 ± 0.02 K km s−1 (Fig. 3). This corresponds to ing the “qphot” task of the “digiphot.apphot” package of IRAF3 ,
a molecular gas mass of MH2 = 7.1 × 105 M . The CO(1–0) 3
IRAF is distributed by the National Optical Astronomy
emission appears to be maximal at the offset (−22 ; 0 ), where Observatories, which are operated by the Association of Universities
it is detected at 3−4σ (Table 2). The CO(2–1) emission is, on for Research in Astronomy, Inc., under cooperative agreement with the
the other hand, not detected down to an integrated line flux National Science Foundation.
A147, page 6 of 12
Fig. 4. Multiwavelength views of the bright UV region at rgal = 1.36 r25 , showing from left to right the NUV, 24 μm, and H I 21 cm emission in
flux units of 10−9 MJy sr−1 , MJy sr−1 , and K km s−1 , respectively, as indicated by the color bars. The circles of 22 diameter refer to the positions
of the 12 pointings used to map the CO(1–0) emission. The Hα emission is undetected in this external region of the M 63 galaxy.
rgal [kpc]
0 5 10 15 20 25 30
CO(2-1) (HERACLES)
1.0
FUV
NUV
HI
0.8
Hα
24μm
UV reg CO(1-0) (this work)
Normalized flux
0.6
0.4
0.2
CO(1-0) x 40 in UV reg
24μm x 200 in UV reg
0.0
0 100 200 300 400 500 600 700

rgal [arcsec]
Fig. 5. Radial profiles of FUV, NUV, Hα, 24 μm, different star formation tracers, and H I taken from the literature (see Table 3), as measured along
the major axis of the M 63 disk, which we compare to our CO(1–0) radial profile (red stars) and to the CO(2–1) radial profile from the HERACLES
survey (Leroy et al. 2009). The measured fluxes are plotted in a normalized flux scale in order to allow relative comparisons. All the star formation
tracers nicely follow the spiral arm structure at rgal 130, 200, and probably 260 . We observe a severe drop with the galactocentric radius for all
star formation tracers, as well as for CO, such that close to the r25 limit (dashed line) and beyond all tracers and CO are practically absent with the
exception of UV and H I. The CO emission is again detected in the external bright UV region. The measurements obtained in this UV region are
shown in the dashed box at the correct galactocentric radii. The CO(1–0) and 24 μm fluxes are multiplied by a factor of 40 and 200, respectively,
to make them visible. The double data points plotted at each rgal correspond to the parallel pointings used to map the UV region (see Table 2). This
external UV region has relatively high FUV and NUV emissions and is dominated by H I.
and we subtracted an average sky value measured away from the profile of the atomic gas, which shows a completely different be-
galaxy. The photometric measurements were performed at the havior. We observe that close to the r25 limit, the star formation
27 positions of the radial cut along the M 63 major axis used for tracers and CO begin to vanish considerably, and beyond the r25
the CO radial mapping, as well as at the 12 positions used to map limit, they all are practically absent with the exception of the
CO in the external bright UV region, following the same spacing faint UV emission. The H I emission is very strong beyond the
and pointings exactly. r25 limit. This shows the importance of looking for molecular
The resulting CO(1–0), CO(2–1), FUV, NUV, H I, Hα, and gas beyond the optical limit, where evidence of star formation is
24 μm radial profiles along the M 63 radial cut are shown in clearly brought by the UV emission.
Fig. 5 (connected data points), where the measured fluxes are While the CO emission is not detected beyond the r25 limit
plotted in a normalized flux scale to allow relative comparisons. in the M 63 major axis map, it is again detected in the external
All the profiles nicely follow the spiral structure observed at UV region at rgal = 1.36 r25. The corresponding CO(1–0) fluxes,
galactocentric radii rgal 130, 200, and probably 260 . They multiplied by a factor of 40 to make them visible (showing the
all show a severe drop with the galactocentric radius, except the depth of our data), with H I and other star formation tracer fluxes
A147, page 7 of 12
A&A 566, A147 (2014)
as measured in the UV region, are also shown in Fig. 5. The rgal [kpc]
log SFE [yr−1] = log ΣSFR/ΣH2

0 5 10 15 20 25 30
FUV and NUV emissions are observed in all pointings over the −8.0
external UV region, while the Hα emission is undetected and a −8.5 Radial cut
faint 24 μm emission, whose fluxes were multiplied by a factor −9.0 UV reg
of 200 to make them visible, is observed in only eight pointings −9.5
(see also Fig. 4). It is interesting to point out that the FUV and −10.0
NUV fluxes in the external UV region are similar to those at −10.5
the pointing at r25 , the galactocentric limit where CO is still de- 0 100 200 300 400 500 600
tected along the M 63 major axis. Beyond r25 , the FUV and NUV rgal [arcsec]
fluxes decrease and get weaker than in the external UV region. rgal [kpc]
0 5 10 15 20 25 30
This suggests a tight correlation between the UV flux, namely 1.5
the intensity of star formation and the CO flux that should trace 1.0 Radial cut
log ΣH2 /ΣH I

the amount of molecular gas, since the CO emission starts to be 0.5 UV reg
detected at a given UV flux level. If correct, the CO emission 0.0
is not detected beyond r25 along the major axis mapping sim- −0.5
ply because the amount of CO, corresponding to the molecular −1.0
gas needed to sustain the star formation that is taking place there, 0 100 200 300 400 500 600
falls below our detection threshold. Consequently, it is likely that rgal [arcsec]
molecular gas is present in the outer regions of the M 63 disk at Fig. 6. Radial profiles of the star formation efficiency (upper panel)
least as long as the XUV is present. and the H2 -to-H I surface density ratio (bottom panel). The SFE is de-
The considerably larger amount of H I observed in the ex- fined here as the star formation rate per unit of molecular gas. The blue
ternal UV region relative to CO leads us to speculate that H I is squares correspond to the 17 pointings used to map CO along the M 63
more likely a precursor of H2 rather than a product of photodis- major axis (the radial cut) out to the isophotal radius, r25 (dashed line),
sociation due to UV radiation. However, the situation might be and the red circles correspond to the 8 pointings with 24 μm emission
more complicated in reality. The main complication we may in- detection used to map CO over the bright UV region at rgal = 1.36 r25 .
voke is the reliability of CO as a proxy of H2 . As discussed in The double data points plotted for the UV region at each rgal correspond
Sect. 4 and shown in Fig. 5, the absence of a clear CO(2–1) line to the parallel pointings used to map this region (see Table 2).
detection beyond rgal = 0.68 r25 , with a CO luminosity ratio, r2,1 ,
below ∼0.1 (see Table 1), may betray a decrease in the excitation
temperature with the galactocentric radius. If this is the case, it galactocentric radius, as shown by Sandstrom et al. (2013, see
may be that even if substantial H2 is present in the outer regions their Fig. 7) for M 63 over the radial cut out to rgal ∼ 0.7 r25 , is
of the M 63 disk, the CO lines may be quite weak, hence remain confirmed.
undetected.
In Fig. 6 we plot the radial profiles of the star formation effi- 5.2. Kennicutt-Schmidt relations
ciency (SFE) and the H2 -to-H I surface density ratio. The SFE is
the star formation rate (SFR) per unit of molecular gas or the in- Our acquired CO measurements offer the opportunity to analyze
verse of the gas depletion timescale, i.e., the time required for the Kennicutt-Schmidt relations across the optical disk of the
present-day star formation to consume the gas reservoir. It is M 63 galaxy and, in particular, close to and beyond the isopho-
calculated here as SFE = ΣSFR /ΣH2 . The various surface den- tal radius, r25 . This exercise is especially of interest in the bright
sities are computed using Eqs. (1) to (3). In the radial cut we UV region located in the outer disk of M 63 at the galactocentric
only consider the pointings up to the isophotal radius, r25 , where radius rgal = 1.36 r25 , whose stellar emission is solely dominated
we obtained convincing CO detections (17 pointings), and in by the UV emission, so where the star formation rate, metallic-
the UV region we retain pointings with 24 μm detections only ity, gas temperature, excitation, and gas density are potentially
(8 pointings) in order to accurately estimate the star formation lower. For this purpose, we use the photometric measurements
rate density. We observe that the SFE, as measured over the ra- obtained in Sect. 5.1 for the FUV, 24 μm, and H I data. At each
dial cut of M 63, is roughly constant as a function of the galacto- pointing of the CO radial mapping and the CO mapping of the
centric radius out to the r25 limit. This is in line with the findings bright UV region, we compute the star formation rate surface
by Leroy et al. (2008), which we now extend out to the isophotal density, ΣSFR , the atomic gas surface density, ΣH I , and the molec-
radius. Beyond r25 , the two SFE detections tend to show a drop ular gas surface density, ΣH2 . We can question how these surface
in the SFE in the bright UV region, but upper limits in the other densities determined over single pointings along the radial major
data points prevent us from drawing definitive conclusions. The axis cut are representative of surface densities that Leroy et al.
H2 -to-H I surface density ratio shows a smooth decrease with (2008) obtained by performing azimuthal averages over the en-
the galactocentric radius, such that the inner regions of the M 63 tire disk of M 63. The comparison of the respective molecular
disk are H2 -dominated, while the regions at rgal > 0.5 r25 end gas surface densities as a function of the galactocentric radius is
up to be H I-dominated. This transition between a “mostly-H2” excellent. The SFR surface densities for the radial cut show an
and a “mostly-H I” interstellar medium (ISM) is found to be offset toward higher values, but the respective ΣSFR radial pro-
a well-defined function of local conditions according to Leroy files are the same. Only the atomic gas surface density measure-
et al. (2008), occurring at characteristic galactocentric radius, ments seem to severely diverge between the radial cut and the
stellar and gas surface densities, hydrostatic gas pressure, and azimuthal averages. This difference comes from the fact that the
orbital timescale. At the r25 limit and beyond, we observe a fur- radial cut along the major axis crosses a spiral arm like structure,
ther and more severe drop of the H2 -to-H I surface density ratio. which is very bright in H I and covers an H I emission peak in the
Nevertheless, the behavior of the radial profiles of the SFE and outer parts of the radial profile. This certainly also explains the
the H2 -to-H I surface density ratio may well be inverted, if the higher ΣSFR we observe for the radial cut, where the star forma-
trend toward an increasing CO-to-H2 conversion factor with the tion activity is slightly enhanced.
A147, page 8 of 12
Table 4. Best-fitting bisector linear Kennicutt-Schmidt relations obtained for M 63 in log space.
This work Bigiel et al. 2008

Radial cut out to r25 UV region at 1.36 r25 Radial cut out to 0.68 r25
log A N σ log A N σ log A N σ
H2 –1.93 1.06 0.29 0.17† 3.0† – –2.22 0.92 0.10
H I + H2 –2.43 1.73 0.31 –2.20 4.7 0.35 –2.63 1.58 0.22
Notes. The Kennicutt-Schmidt relations are best parametrized by power laws of the form ΣSFR = A(Σgas )N , which in log space translates into linear
relations. † Only indicative log A and N values, since they rely only on two data points, which makes them very uncertain.
The SFR surface density is determined using the calibration made this rescaling to minimize the covariance of log A and N
from Leroy et al. (2008): in the fit4 and to make our results comparable to those of Bigiel
et al. (2008) for M 63. They used the same FUV and H I data
ΣSFR (M yr−1 kpc−2 ) = as we did, but the CO data were taken from the HERACLES
CO(2–1) database (Leroy et al. 2009) with a 11 resolution and
8.1 × 10−2 FFUV (MJy sr−1 ) + 3.2 × 10−3 F24 μm (MJy sr−1 ), (1)
a CO(2–1)/CO(1–0) line ratio of 0.8. Our CO(1–0) data reach
which includes the FUV flux to measure the unobscured star for- much deeper sensitivity than the HERACLES CO(2–1) data,
mation through the emission of O and B stars, and the 24 μm flux which is essential for properly characterizing the K-S relations
that traces the obscured FUV emission that is re-emitted in the over all the optical disk and especially beyond the r25 limit.
far-IR by dust grains. The atomic gas surface density is measured The best-fitting bisector linear relations, log ΣSFR = log A +
with the calibration of Bigiel et al. (2010): N log Σgas , obtained independently for the K-S relation over the
full radial cut out to r25 and the UV region in the outer disk of
ΣH I (M pc−2 ) = 0.020FH I (K km s−1 ), (2) M 63 are shown in Fig. 7. The errors we provide on log A and N
are 1σ quotes of the nearly Gaussian distributions of log A and N
which includes a factor of 1.36 to reflect the presence of helium. obtained through Monte-Carlo simulations of 5000 data sets cre-
Finally, the molecular gas surface density (see Tables 1 and 2) ated from the original data, with random values generated within
is calculated with our integrated CO(1–0) line fluxes using the their error bars. The scatter in our fits typically are on the order of
calibration of Leroy et al. (2008): 0.3 dex. The corresponding K-S relation results are presented in
ΣH2 (M pc−2 ) = 4.4 cos(i)FCO(1−0) (K km s−1 ), (3) Table 4 for both Σgas = ΣH2 and Σgas = ΣH I+H2 , and are compared
to the respective K-S relations found by Bigiel et al. (2008) over
where the “standard” Galactic CO-to-H2 conversion factor, the radial cut of the M 63 disk out to rgal = 0.68 r25 only. The
XCO = 2×1020 cm−2 (K km s−1 )−1 (Dickman et al. 1986), as well K-S relations obtained over the radial cut have very comparable
as the factor of 1.36 to account for helium are adopted. The incli- slopes (in log space) N ∼ 1.0 for ΣH2 and N ∼ 1.6 for ΣH I+H2 ,
nation correction by an angle i = 55◦ is taken into account. This with our slopes systematically slightly steeper. They also have
correction is also considered in the SFR and atomic gas surface comparable coefficients log A ∼ −2.0 for ΣH2 and log A ∼ −2.5
densities when deprojecting the FUV, 24 μm, and H I images for ΣH I+H2 , with our coefficients systematically smaller by about
(see Sect. 5.1). 0.2−0.3 dex, but still within the fit scatter. The one-to-one lin-
We compare the star formation rate surface densities with earity of the K-S relation between the SFR surface density and
the atomic and molecular gas surface densities separately in the the molecular gas surface density is really solid, meaning that
two top panels of Fig. 7. The two plots show that the gas sur- the gas is being consumed at a nearly constant rate, as expected
face density of M 63 in its radial cut is dominated by the molec- for ΣH2 .
ular gas, whereas the bright UV region is strongly dominated When comparing these results to the K-S relations of the
by the atomic gas, as also observed in Fig. 6 (bottom panel). UV region in the outer disk of M 63 at the galactocentric ra-
Interestingly, the H I surface density of the UV region is very dius rgal = 1.36 r25, we note that they are significantly differ-
similar to the one observed along the radial cut, in contrast to ent. The slope N = 4.7 (in log space) of the K-S relation for
its molecular gas surface density, which is significantly lower. ΣH I+H2 is considerably steeper in the external UV region. A sim-
Finally, in the bottom panel of Fig. 7 we show the star formation ilar trend is observed for ΣH2 . It is true that this relies on two
rate surface density as a function of the total gas surface density, data points, but if we trust our 3σ upper limits on ΣH2 , we can
ΣH I+H2 , namely the sum of both ΣH I and ΣH2 . hardly expect to reconcile these external UV region measure-
To characterize the star formation in M 63, we use the ments with the K-S relations observed over the radial cut, more
Kennicutt-Schmidt (K-S) relation that relates the SFR surface especially as the derived star formation rate surface densities are
density to the gas surface density. In all plots of Fig. 7, we fit very reliable. It is very likely that the CO-to-H2 conversion fac-
a power law of the form ΣSFR = A(Σgas )N , which in log space tor adopted over the external UV region is underestimated, given
translates into a simple linear relation. The coefficient A traces the low metallicity expected in the outskirts of M 63 as supported
the absolute star formation efficiency (Kennicutt 1998), and the by the metallicity gradient inferred in M 63 (see Sect. 2) and the
exponent N relates the star formation rate to the gas density observed trend for an increase in XCO with decreasing metal-
present and tells us how close the two variables are to linearity. licity (Sandstrom et al. 2013; Bolatto et al. 2013). While this
We use the least-square bisector method, which is suitable for may induce a shift in the external UV region data points toward
two independent variables (Isobe et al. 1990), for the fit in log higher gas surface densities, the slopes of the K-S relations for
space and leave log A and N as free parameters. It is important
to note that for the fitting procedure alone (not the plotting), we 4
This rescaling has no influence on the results of the exponent N, so
rescaled Σgas by a factor of 10, such that Σgas /10 M pc−2 . We only the coefficient A will need to be corrected by the factor of 10.
A147, page 9 of 12
A&A 566, A147 (2014)
0 0
−1 −1
−2 −2
log(ΣSFR) [M yr−1 kpc−2]

100% 100%
−3 −3
10% 10%
−4 −4
1% 1%
−5 −5
radial cut: log A=-1.23±0.05, N=2.85±0.15, σ=0.78 radial cut: log A=-1.93±0.01, N=1.06±0.01, σ=0.29
UV reg: log A=-2.13±0.22, N=4.42±0.52, σ=0.35 UV reg: log A=0.17±3.23, N=3.0±2.48 (indicative)
−6 −6
−1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5 −1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5
log(ΣHI) [M pc−2] log(ΣH2 ) [M pc−2]
−1
−2
100%
−3
10%
−4
1%
−5
Radial cut, log A=-2.43±0.01, N=1.73±0.01, σ =0.31
UV region, log A=-2.2±0.23, N=4.7±0.6, σ =0.35
−1.0 −0.5 0.0 0.5 1.0 1.5 2.0 2.5
log(ΣHI+H2 ) [M pc−2]
Fig. 7. Kennicutt-Schmidt relations that relate the SFR surface density to the gas surface density plotted for ΣH I (top-left panel), ΣH2 (top-right
panel), and ΣH I+H2 (bottom panel). The blue squares correspond to the 17 pointings used to map CO along the M 63 major axis (the radial cut)
out to the isophotal radius, r25 , and the red circles correspond to the 8 pointings with 24 μm emission detection used to map CO over the bright
UV region at rgal = 1.36 r25 . The K-S relation is best parametrized by a power law of the form ΣSFR = A(Σgas )N , which in log space translates
into a linear relation. The best-fitting bisector linear K-S relations are derived separately for the radial cut (blue dashed-dotted lines) and for the
UV region (red dashed lines). The corresponding coefficients log A and slopes N can be found in the labels of each panel. The black dotted lines
represent “isochrones” of constant star formation efficiencies, indicating the level of ΣSFR needed to consume 100%, 10%, and 1% of the total
amount of gas within 108 years. The vertical dashed line in the top-left panel corresponds to the ΣH I ∼ 9 M pc−2 threshold at which the atomic
gas saturates (Bigiel et al. 2008).
ΣH I+H2 and ΣH2 will remain unchanged, unless there is a sig- of star formation drastically changes beyond the isophotal ra-
nificant XCO gradient over the external UV region. As a result, a dius, r25 . Indeed, the exponents N observed for the K-S rela-
broken K-S power law between the inner and outer regions of the tions in the external UV region show a nonlinear SFR regime, in
M 63 disk seems to robustly emerge. It suggests that the regime clear contrast to the quasi-linear SFR regime in the inner regions
A147, page 10 of 12
of the disk. What triggers the “quenching” of star formation in the fourth molecular gas detection in the outskirts of nearby
the outer regions of the M 63 disk and, consequently, the severe spiral galaxies. The CO(2–1) emission is, on the other hand,
deviation from the commonly accepted star formation relation confined to rgal = 0.68 r25 and thus suggests subthermal ex-
as traced by the K-S relations observed along the radial cut of citation in the outer regions of the M 63 disk.
M 63? We can invoke the flaring of the outer gas layers or, more 2. The radial profiles of the CO emission and other star for-
speculative, the possible presence of high turbulence providing mation tracers (Fig. 5) show a severe drop with the galac-
support against gravitational collapse as suggested by Longmore tocentric radius, in contrast to the atomic gas. Close to the
et al. (2013). r25 limit, CO and the star formation tracers begin to vanish
To better appreciate the change in the star formation regime considerably, and beyond r25 , they all are practically absent
between the inner and outer regions of the M 63 disk, in Fig. 7 with the exception of the faint UV emission and H I. The
we also plot “isochrones” of constant star formation efficiencies, UV region at rgal = 1.36 r25 , in which the CO emission is
indicating the levels of ΣSFR needed to consume 100%, 10%, and detected, is characterized by FUV and NUV emission fluxes
1% of the total amount of gas (corrected by a factor of 1.36 to similar to the fluxes observed at r25 , but stronger than the
account for helium) within 108 years. These isochrones can also fluxes observed at rgal > r25 along the M 63 major axis. This
be interpreted as constant gas depletion timescales (i.e., the time probably reflects a tight correlation between the CO and UV
needed to consume the total amount of gas) of 108 yr, 109 yr, fluxes, namely between the intensity of star formation and
and 1010 yr from top to bottom. We can see that when the total the amount of molecular gas, so it strongly suggests that the
gas is considered (H I + H2 ), the SFE settles the gas consumption absence of CO detection at rgal > r25 , where the XUV is
within 108 years to 1% to 10% for the radial cut of M 63, whereas weaker, is simply the result of the CO detection threshold
the external UV region has a much lower SFE with less than that is still too high.
1% of the gas converted into stars within 108 years. The drop 3. The external UV region is characterized by a very high
in the star formation efficiency beyond the isophotal radius, r25 , H I flux with respect to the measured CO flux. This leads us
can also be appreciated in Fig. 6 (upper panel). The SFE could to speculate that H I is more likely the precursor of H2 rather
even be lower in the external UV region, since the CO-to-H2 than the product of UV photodissociation, since it seems to
conversion factor may, in reality, be up to 100 times more (in the dominate in quantity. This is, however, true as long as sub-
extreme case) than the “Galactic” value (Bolatto et al. 2013) and stantial H2 is not hidden in the outer disk regions of M 63,
so are the computed ΣH2 values, depending on the metallicity of which may be the case as we observe hints for an excita-
the external UV region, which may well be as low as 10% to 20% tion temperature decrease at large galactocentric radii, which
of the local Galactic ISM5 . Consequently, the star formation still may lead to very weak CO lines.
occurs in the outer regions of the M 63 disk, but at very low 4. With the integrated CO line flux measurements and the com-
efficiency, significantly lower than in the inner regions of the plementary data from the literature, we derive SFR, H I, and
disk. H2 surface densities all along the major axis of the M 63
disk and in the external UV region. We observe that the
gas surface density along the radial cut is dominated by the
6. Summary and conclusions molecular gas, whereas in the UV region it is dominated
by the atomic gas. The best-fit Kennicutt-Schmidt relations,
Deep CO(1–0) and CO(2–1) observations obtained on the IRAM ΣSFR = A(Σgas )N , show a broken power law from the inner to
30 m telescope of the M 63 spiral galaxy characterized by an the outer regions of the M 63 disk (Fig. 7). Indeed, the almost
XUV disk extending out to 2.5 times the optical equivalent ra- linear K-S relation (with a slope of nearly 1 in log space) ob-
dius were presented. We performed both a CO mapping along served over the radial cut, in the inner regions of the disk
the major axis of the M 63 disk from the center out to the galac- characterized by high gas densities, cannot be extrapolated
tocentric radius rgal = 572 = 1.6 r25 and over a bright UV to the outer disk regions. The latter are characterized by a
region in the outer disk of M 63 at the galactocentric radius nonlinear SFR regime (with a K-S slope much higher than 1
rgal = 1.36 r25. Our objective was to search for CO emission in log space), perhaps owing to the flaring of the outer gas
and hence for molecular gas in the outer regions of the M 63 layers. This is the first time that the K-S relation is quanti-
disk beyond the optical radius, where evidence of star formation fied in the outskirts of a spiral galaxy, i.e., in low gas density
is brought by both the extended UV emission and high H I col- environments. At a molecular gas surface density as low as
umn densities observed in these regions. We highlighted the im- ΣH2 = 0.35 M pc−2 , well below all the determined H2 sur-
portance of a CO detection in regions far away from the center face densities referenced in spiral galaxies so far (Bigiel et al.
of the galaxy, where the metallicity, gas temperature, excitation, 2008, 2011), star formation still occurs spontaneously.
and gas density are supposed to be lower and where stars have 5. The change in the star formation regime between the inner
more difficulty to form. To complement our CO observations, and outer regions of the M 63 disk can also be appreciated
we used FUV, NUV, Hα, 24 μm, and H I data from the literature. by the difference in their star formation efficiencies. Indeed,
This allowed us to investigate the Kennicutt-Schmidt relations along the major axis of the disk out to the isophotal radius
across the galaxy and beyond the isophotal radius, r25 , in the the SFE settles the gas consumption within 108 years from
bright UV region. Our main results are as follows. 1% to 10%, whereas in the external UV region much less
1. The CO(1–0) emission is clearly detected along the major than 1% of the gas is converted into stars within 108 years.
axis of the M 63 disk out to the isophotal radius, r25 , but not Consequently, star formation still occurs in the outer regions
beyond. However, the CO(1–0) is again detected in the bright of the disk, but at very low efficiency.
UV region in the outer disk of M 63 at rgal = 1.36 r25. This is
Acknowledgements. C.V. wishes to acknowledge support from CNRS and
5
When extrapolating the metallicity gradient determined for M 63 (see CONICYT through an agreement signed on December 11, 2007. We warmly
Sect. 2), we get a metallicity of 20% of the local Galactic ISM at rgal = thank the IRAM 30 m telescope staff for their support during the observations.
1.36 r25 , the galactocentric radius of the bright UV region. We thank the anonymous referee for her/his very careful and constructive report.
A147, page 11 of 12
A&A 566, A147 (2014)
References Gil de Paz, A., Boissier, S., Madore, B. F., et al. 2007, ApJS, 173, 185
Helfer, T. T., Thornley, M. D., Regan, M. W., et al. 2003, ApJS, 145, 259
Alberts, S., Calzetti, D., Dong, H., et al. 2011, ApJ, 731, 28 Henry, R. B. C., & Worthey, G. 1999, PASP, 111, 919
Allen, R. J. 1996, Cold dust and Galaxy Morphology, eds. D. L. Block, & J. Isobe, T., Feigelson, E. D., Akritas, M. G., & Babu, G. J. 1990, ApJ, 364, 104
M. Greenberg, Astrophys. Space Sci. Lib. (Dordrecht: Kluwer Academic Kennicutt, Jr., R. C. 1989, ApJ, 344, 685
Publishers), 209, 50 Kennicutt, Jr., R. C. 1998, ApJ, 498, 541
Allen, R. J., Atherton, P. D., & Tilanus, R. P. J. 1986, Nature, 319, 296 Kennicutt, Jr., R. C., Lee, J. C., Funes, S. J., et al. 2008, ApJS, 178, 247
Allen, R. J., Heaton, H. I., & Kaufman, M. J. 2004, ApJ, 608, 314 Leroy, A. K., Walter, F., Brinks, E., et al. 2008, AJ, 136, 2782
Battaglia, G., Fraternali, F., Oosterloo, T., & Sancisi, R. 2006, A&A, 447, 49 Leroy, A. K., Walter, F., Bigiel, F., et al. 2009, AJ, 137, 4670
Bigiel, F., Leroy, A., Walter, F., et al. 2008, AJ, 136, 2846 Longmore, S. N., Bally, J., Testi, L., et al. 2013, MNRAS, 429, 987
Bigiel, F., Leroy, A., Walter, F., et al. 2010, AJ, 140, 1194 Lupton, R., Blanton, M. R., Fekete, G., et al. 2004, PASP, 116, 133
Bigiel, F., Leroy, A. K., Walter, F., et al. 2011, ApJ, 730, L13 Martin, C. L., & Kennicutt, Jr., R. C. 2001, ApJ, 555, 301
Bolatto, A. D., Wolfire, M., & Leroy, A. K. 2013, ARA&A, 51, 207 Moustakas, J., Kennicutt, Jr., R. C., Tremonti, C. A., et al. 2010, ApJS, 190,
Braine, J., & Herpin, F. 2004, Nature, 432, 369 233
Braine, J., Ferguson, A. M. N., Bertoldi, F., & Wilson, C. D. 2007, ApJ, 669, Nieten, C., Neininger, N., Guélin, M., et al. 2006, A&A, 453, 459
L73 Pilyugin, L. S., & Thuan, T. X. 2005, ApJ, 631, 231
Braine, J., Gratier, P., Kramer, C., et al. 2010, A&A, 520, A107 Sancisi, R., Fraternali, F., Oosterloo, T., & van der Hulst, T. 2008, A&ARv, 15,
Combes, F., & Pineau Des Forets, G. 2000, Molecular Hydrogen in Space 189
(Cambridge: Cambridge University Press) Sandstrom, K. M., Leroy, A. K., Walter, F., et al. 2013, ApJ, 777, 5
Crosthwaite, L. P., Turner, J. L., Buchholz, L., Ho, P. T. P., & Martin, R. N. 2002, Schruba, A., Leroy, A. K., Walter, F., et al. 2011, AJ, 142, 37
AJ, 123, 1892 Smith, D. A., Allen, R. J., Bohlin, R. C., Nicholson, N., & Stecher, T. P. 2000,
Cuillandre, J.-C., Lequeux, J., Allen, R. J., Mellier, Y., & Bertin, E. 2001, ApJ, ApJ, 538, 608
554, 190 Solomon, P. M., Rivolo, A. R., Barrett, J., & Yahil, A. 1987, ApJ, 319, 730
Dale, D. A., Cohen, S. A., Johnson, L. C., et al. 2009, ApJ, 703, 517 Solomon, P. M., Downes, D., Radford, S. J. E., & Barrett, J. W. 1997, ApJ, 478,
de Blok, W. J. G., & Walter, F. 2003, MNRAS, 341, L39 144
Dickman, R. L., Snell, R. L., & Schloerb, F. P. 1986, ApJ, 309, 326 Thilker, D. A., Bianchi, L., Boissier, S., et al. 2005, ApJ, 619, L79
Dong, H., Calzetti, D., Regan, M., et al. 2008, AJ, 136, 479 Walter, F., Brinks, E., de Blok, W. J. G., et al. 2008, AJ, 136, 2563
Ferguson, A. M. N., Wyse, R. F. G., Gallagher, J. S., & Hunter, D. A. 1998, ApJ, Young, J. S., & Scoville, N. 1982, ApJ, 258, 467
506, L19 Young, J. S., & Scoville, N. Z. 1991, ARA&A, 29, 581
Gil de Paz, A., Madore, B. F., Boissier, S., et al. 2005, ApJ, 627, L29 Young, J. S., Xie, S., Tacconi, L., et al. 1995, ApJS, 98, 219
A147, page 12 of 12
2.2 2013 observations
Motivated by the successful results obtained in M63, new CO observations were
done in the external parts of several XUV disk galaxies. Such galaxies were selected
for having large disks of UV emission, beyond their r25 limit, previous CO detection
in their centers (Holwerda et al. 2012, Young et al. 1995), and a low inclination
angle to avoid projection problems. With this criteria, three spiral galaxies were
selected: NGC628, NGC3344 and NGC2403, choosing the brightest points in FUV-
NUV as targets to be observed in CO(1–0) and CO(2–1) (Figure 2.13 and Table 2.8).
Details of these galaxies, including positions, distances, r25 limits and integration
times during the observations are presented in Table 2.7.
For this project a total of 40 hours were granted between January 31th and
February 4th 2013. Atmospheric conditions were good (pwv between 1.2 and 2.6
mm), but due to strong winds ∼40% of the time was lost. A total of 26.6 hrs were
observed, including overheads.
2.2.1 Data reduction and Results
The data were recorded with the WILMA and FTS correlators, which give a
spectral resolution of 2 MHz and 195 kHz respectively. At 115 GHz, these resolutions
translate into 5.2 km/s for WILMA and 0.5 km/s for FTS, and at 230 GHz, the
corresponding values are 2.6 km/s and 0.25 km/s respectively.
The data reduction was done with CLASS, in the same way as the reductions of
M63. Each spectra of both backend were carefully inspected, and bad scans were
removed.
Baselines were subtracted with polynomials of order 0 and 1 depending on the
source, and antenna temperatures were corrected by the telescope beam efficiency
(Beff ) and forward efficiency (Feff ) to obtain main beam temperatures with Beff and
Feff equal to 0.78 and 0.94 respectively for 115 GHz, and to 0.59 and 0.92 for 230
GHz.6
Spectra were smoothed with the hanning method to degrade the velocity resolu-
tion until obtain a value no greater than 1/3 of the FWHM line. Finally, a simple
gaussian line was fitted to the line candidate. The CLASS fit returns the veloc-
ity position of the line, its FWHM, the peak temperature and the integrated line
intensity.
Unfortunately, and probably due to the bad weather conditions, no CO detections
were found beyond the r25 limit of each galaxy. Apart from confirming the CO
detections in the centers of NGC2403 and NGC3344, only one detection was found
in the point NGC3344-5 located at 2.6 kpc from the center of the galaxy (Figure
2.14). Details of line parameters of such detections are presented in Table 2.9.
6
http://www.iram.es/IRAMES/mainWiki/Iram30mEfficiencies
63
NGC628-1 NGC3344-2
NGC628-2
NGC3344-1
NGC628-3
NGC3344-5
NGC3344-3
NGC3344-4
NGC628-4
NGC2403-1
NGC2403-2
NGC2403-4 NGC2403-3
NGC2403-5
NGC2403-6
Figure 2.13: Targets selected for NGC628 (top left), NGC3344 (top right) and NGC2403 (bottom)
in GALEX FUV-NUV composite images (FUV in red and NUV in blue, Gil de Paz et al. 2007a).
The green ellipse marks the r25 limit. Circles are 2200 in diameter, as the CO(1–0) FWHM.
Observations were done as single pointing observation in each one of these targets.
Table 2.7. Galaxies selected for XUV observations in 2013.

Galaxy RA(J2000)a DEC(J2000)a da r25 a int. time
hr m s
[ : : ] deg m
[ : : ]s [Mpc] [arcmin] [hrs]
NGC628 01:36:41.8 15:47:00.5 11 5.0±0.1 ···

NGC2403 07:36:51.4 65:36:09.2 3.2 10.0±0.1 0.7
NGC3344 10:43:31.2 24:55:20.0 6.9 3.4±0.1 10.9
a Information taken from Gil de Paz et al. (2007a).
Note. — Due to bad weather conditions NGC628 could not be observed.
64
Table 2.8. Targets description for XUV observations 2013.
Target RA(J2000) DEC(J2000) int. time detection
[hr :m :s ] [deg :m :s ] [hrs]
NGC2403-1 07:35:31.42 65:38:40.78 0.4 no

NGC3344-1 10:43:31.80 24:57:57.58 1.6 no
NGC3344-2 10:43:33.62 24:59:13.33 0.9 no
NGC3344-3 10:43:15.59 24:55:35.02 4.0 no
NGC3344-4 10:43:19.02 24:53:28.29 3.3 no
NGC3344-5 10:43:30.81 24:56:35.00 0.6 yes
Table 2.9. Line parameters for detections in XUV observations 2013.

Source line resol. v0 dv Tmb rms ICO M2
[km/s] [km/s] [km/s] [mK] [mK] [K km/s] [×106 M ]
NGC2403 CO(1–0) 4 126.0±1.5 40.3±3.5 60 10 2.58±0.19 1.04

··· CO(2–1) 2 1425.4±1.0 32.4±2.3 48 8.4 1.65±0.10 ···
NGC3344 CO(1–0) 4 583.7±2.4 64.1±5.8 32 7.3 2.18±0.17 4.08
··· CO(2–1) 4 1878.6±0.9 56.9±2.5 72 6.4 4.39±0.15 ···
NGC3344-5 CO(1–0) 4 643.6±0.7 20.2±1.5 55 5.6 1.18±0.08 2.2
··· CO(2–1) 4 1939.3±1.1 23.3±2.3 37 5.8 0.92±0.08 ···
Note. — Data for NGC2403 and NGC3344 correspond to their galactic centers. H2 masses were cal-
culated with Equation 1.4, i e. MH2 [M ] = 4.4πR2 [pc]ICO(1−0) [K km/s], with the radius R measured
as the CO(1–0) beamsize radius (1100 ) at the distances of the galaxy, from Table 2.7.
65
Figure 2.14: Spectra results for XUV observations 2013.
66
Table 2.10. 3σ upper limits of CO(1–0) emission for sources with no detection in
2013 observations.
Source rms ICO MH2
[mK] [K km/s] [×105 M ]
NGC2403-1 3.4 <0.31 <1.25

NGC3344-1 1.6 <0.14 <2.62
NGC3344-2 1.9 <0.17 <3.18
NGC3344-3 1.1 <0.10 <1.87
NGC3344-4 1.2 <0.11 <2.06
Note. — Rms values correspond to Tmb scale,

and were calculated in smoothed spectra of ∼30 km/s
resolution. ICO upper limits calculated assuming a
velocity line width of 30 km/s.
For the remaining sources though, we calculated 3σ upper limits for the CO(1–0)
emission, from the rms values of the spectra, and assuming a velocity width of 30
km/s. Such limits, along with their corresponding mass estimation, are listed in
Table 2.10
2.2.2 Stacking
In regions where the CO is hard to detect due to its faintness, it is not surprising
that no detection was found beyond the r25 limits, especially in bad weather condi-
tions. But we cannot assure that the CO is not present. All we can say is that the
S/N is too low to confirm a detection.
One way to overpass this problem is through stacking of the CO spectra. By
averaging the spectra within a given region, we can increase the S/N and verify that
faint emission is in fact an astronomical signal (Schruba et al. 2011). It is important
to note that when stacking the spectra of different targets in the same galaxy we are
loosing position information, but gaining crucial S/N. If a detection appears after
the stacking of several spectra obtained at different positions, we can affirm that the
detection is an average value inside a given range of radii, rmin and rmax , where the
targets are located.
Now, in order to stack the spectra and average them, first we need to align them
all to the same central velocity. So the most convenient thing to do is to center all the
spectra to v = 0. To do this shifting in velocity we first need to know the velocity at
which we would expect to find the source. Such velocities can be determined through
HI velocity maps, where we can locate the position of our targets and identify the
corresponding velocity. After the shifting in ∆v of each spectra taken in one galaxy,
we average all the spectra centered at v = 0 and obtain one spectra per galaxy.
Figure 2.15 shows this procedure done for NGC3344, were sources 1 through 4 were
67
Figure 2.15: Stacking of CO spectra for NGC3344, sources 1 through 5. The velocity shifting
of the spectra was done by using the HI velocities from Verdes-Montenegro et al. (2000). The
stacked spectra have a resolution of 4 km/s. The rms values of the spectra are 2.82 and 1.85 mK
for CO(1–0) and CO(2–1) respectively.
shifted and averaged all together. Unfortunately no detection was found either after
this technique. This stacking was not done for NGC2403 because only one source
could be observed (with no detection) apart from the galaxy center.
2.3 M83
The last work done in the XUV topic for this thesis was in the nearby spiral
galaxy M83, including exclusive ALMA data from Cycle-2. Due to the very recent
delivery of these data (March 2015), this is a work still in progress.
M83 (NGC 5236) is a remarkable example of a nearby spiral galaxy with an
extended XUV disk reaching 2.3 times the optical major axis radius R25 = 6.090 =
8.64 kpc (Gil de Paz et al. 2007a). Figure 2.18 shows the GALEX NUV and FUV
color-composite image of M 83. An extensive population of UV-bright star-forming
regions (tracing O and B stars) and stellar clusters is revealed. The NUV and FUV
surface-brightness profiles show the smooth decrease of the XUV emission out to 28
kpc in the equivalent radius (a × b)1/2 . Located at a distance of 4.88 Mpc (Paturel
et al. 2003), this SAB(s)c galaxy is expected to be representative of a large class of
spirals, having no immediate neighbours with which it could interact. The 21 cm
observations of M 83 also show the presence of a very large, gaseous disk extending
out to 56 kpc (e.g. Tilanus & Allen 1993). The extent of this HI emission can be
appreciated in Figure 2.18 in red contours. This provides another evidence of the
presence of molecular gas out to large galactocentric radii, since recent studies have
demonstrated that strong H I emission is on average a good tracer of regions rich in
68
Figure 2.16: ALMA interferometer at Llano Chajnantor in Chile. Credit: ALMA
(ESO/NAOJ/NRAO)
molecular gas (e.g. Crosthwaite et al. 2002, Nieten et al. 2006).
2.3.1 Instrumentation and Observations
The Atacama Large Millimeter Array (ALMA) is the largest ground-based astro-
nomical project in existence. Located at 5000 mts above sea level in Llano Chajnan-
tor at the Atacama desert in Chile (Figure 2.16), it is the best interferometer in the
world, in terms of sensitivity and resolution. With its 66 high-precision antennas
spread over distances of up to 16 km, it studies the light coming from the coldest
parts of the universe, at millimeter and sub-millimeter wavelengths, shedding light
on our origins. It is nowadays the most sought-after radiotelescope in the south
hemisphere, and thanks to its unprecedented sensitivity, it is ideal for the search of
molecular gas in the faintest regions of XUV disk galaxies.
With its 8 spectral bands it has a frequency coverage between 84 and 950 GHz,
equivalent to 0.32 and 3.57 mm in wavelength scale (Figure 2.17).
For this project we selected the band 6 to observe CO(2–1) at 229.67 GHz, as
well as 3 continuum bands, centered at 232.5, 218.4 and 216.5 GHz.
For choosing the target for these observations we used the same selection criteria
as for the other XUV galaxies in this chapter, i.e, by analysing the FUV, NUV and
atomic gas maps, we select the regions with peaks of emission in UV and HI, beyond
the optical radius. This is how we selected the region at 13h 37m 03.6s , −29◦ 59m 47.6s
shown in Figure 2.18, enclosed in a green rectangle of 30 x1.50 and located at rgal =
7.850 = 11kpc. Unlike our previous IRAM-30 mt observations, where we were limited
to do single ON+OFF pointings in order to reach the S/N needed in a reasonable
amount of integration time, with ALMA we can take advantage of its much better
sensitivity, and map the entire region within just an hour of observations.
The selected region was observed during ∼1 hour on March 2, 2014, in very good
weather conditions (pwv ∼1.3 mm). The 12m array was used, with 34 antennas and
a maximum baseline of 558.2 m. The map was done with 121 pointings separated
by 12.900 , and with an integration time of 10.8 sec in each one of them.
69
Figure 2.17: Atmospheric transmissions at 0.5, 1.5 and 2.5 mm of pwv for the 8 ALMA bands.
Band 6 was used in this project to observe CO(2–1) at 229.67 GHz.
The initial goal of achieving an rms of 11.1 mJy was reached and surpassed up
to 10.3 mJy, with an angular resolution of 1.600 and a spectral one of 1.94 MHz (2.5
km/s).
2.3.2 Results
The data was reduced using the CASA7 package, version 4.2.2. Approximately
36% of the data was flagged, which is considered normal for 12m Array data.
A CO(2–1) data cube was created with natural weighting, with a velocity range
of ∼15 to ∼1015 km/s, a channel spacing of 2.5 km/s and an rms of 10.3 mJy per
channel. Unfortunately, according to the observing results sent by the ALMA staff,
and after a first inspection of the data cube, no CO(2-1) emission has been found
yet in the selected region in M83 (Figure 2.19). It is possible that no CO(2–1)
emission is present due to a low metallicty and gas density, possibly due to gas
flaring in this region so far away from the center of the galaxy, and so vulnerable to
external factors. This still is a work in progress though, and the are some possible
candidates for clumps that could be real (see Figure 2.20) and will be analyzed in
the near future, with more elaborated algorithms of search for weak emission.
Only one weak point source was detected in continuum, with a flux <2 mJy. This
is a rather unexpected continuum source at such a low frequency, and it is probable
7
the Common Astronomy Software Applications is developed by an international consortium of
scientists based at the National Radio Astronomical Observatory (NRAO), the European South-
ern Observatory (ESO), the National Astronomical Observatory of Japan (NAOJ), the CSIRO
Australia Telescope National Facility (CSIRO/ATNF), and the Netherlands Institute for Radio
Astronomy (ASTRON) under the guidance of NRAO.
70
Figure 2.18: RGB composite color image of M83, with the NUV emission in red, FUV in blue
and a combination of both in green. The arcsinh function has been applied to this image which
allows to show faint objects while simultaneously preserving the structure of brighter objects in
the field, such as the spiral arms of large galaxies (Lupton et al. 2004). The red contours shows
the HI emission taken from the THINGS survey (Walter et al. 2008). The green ellipse shows the
optical limit (r25 = 6.090 = 8.64 kpc). The green rectangle (30 ×1.50 ) marks the region selected to
observe, located at rgal = 7.850 = 11kpc, which is bright in UV and HI emission.
71
M83 CO(2-1)
−29◦ 59′ 00′′ 2.4
1.6
30′′
Jy/Beam Km/s
0.8
DEC. (J2000)
0.0
−30◦ 00′ 00′′
−0.8
−1.6
30′′
−2.4
12s 08s 04s 13h 37m 00s 13h 36m 56s

R.A. (J2000)
Figure 2.19: Intensity map from CO(2–1) data cube. The white circle on the lower left side
indicates the 1.600 beamsize
that a background source, such as a quasar, might be the responsible. This will be
check in the near future of this work. The continuum map (Figure 2.21) has an rms
of 0.19 mJy and synthesized beamsize of 0.7800 ×0.6000 .
2.3.3 Star formation rate
To calculate the SFR taking place in this region, we used complementary data
for NUV and FUV (Gil de Paz et al. 2007a), HI (Walter et al. 2008) and 24µm
emission (Bendo et al. 2012). These maps are presented in Figure 2.22
Same as for M63, we used the FUV and 24µm emission to calculated the SFR
surface density, which includes the FUV flux to measure the unobscured and recent
star formation through the emission of O and B stars, and the 24 µm flux that
traces the obscured FUV emission that is re-emitted in the FIR by dust grains.
From Leroy et al. (2008):
ΣSFR [M yr−1 kpc−2 ] = (8.1×10−2 IFUV [MJy sr−1 ]+3.2×10−3 I24µm [MJy sr−1 ])×cos(i)
(2.4)
where i corresponds to the inclination angle of the galaxy, 24◦ .
We reprojected the FUV map from Figure 2.22 to the same resolution and spatial
grid of the 24µm, and created a ΣSFR map following Eq. 2.4, pixel by pixel. This
72
Figure 2.20: Hint of a possible clump in the CO(2–1) data cube of M83. It will be analyzed
more carefully in the near future, with a special algorithm of search for weak emission.
73
1.8
−29◦ 59′ 00′′ 1.5
1.2
30′′ 0.9
DEC. (J2000)
mJy/Beam
0.6
0.3
−30◦ 00′ 00′′
0.0
−0.3
30′′
−0.6
−0.9
08s 04s 13h 37m 00s
R.A. (J2000)
Figure 2.21: Continuum map of M83 with a weak detection on a point source (<2mJy). The
rms of the map is 0.19 mJy.
×10−2 ×10−2
◦ ′ ′′
−29 59 00 4
4.5
30′′ 3
MJy/sr
MJy/sr
δ2000
3.0
−30◦ 00′ 00′′ 2
1.5
1
30′′
NUV FUV
0
×102
−29◦ 59′ 00′′

1.6
2.4
′′
30
Jy/beam m/s
1.2
MJy/sr
1.6
δ2000
−30◦ 00′ 00′′ 0.8
0.8
30′′ 0.4
HI 24µ
0.0
12s 08s 04s 13h 37m 00s 12s 08s 04s 13h 37m 00s
α2000 α2000
Figure 2.22: Complementary data for M83, with NUV, FUV, HI and 24µm maps.
74
×10−3 ×10−3
0.8 0.8
−29◦ 59′ 00′′ −29◦ 59′ 00′′
0.6 0.6
30′′ 30′′
δ2000
δ2000
0.4 0.4
−30◦ 00′ 00′′ −30◦ 00′ 00′′
0.2 0.2
30′′ 30′′
FUV part FUV part
0.0 0.0
×10−2 ×10−3
2.4
◦ ′ ′′ ◦ ′ ′′
−29 59 00 0.75 −29 59 00
1.6
30′′ 30′′
0.50
δ2000
δ2000
−30◦ 00′ 00′′ −30◦ 00′ 00′′ 0.8
0.25
30′′ 30′′ 0.0
24µ part 24µ part
0.00
×10−2 ×10−3
−29◦ 59′ 00′′ −29◦ 59′ 00′′ 2.4
0.75
30′′ 30′′ 1.6
0.50
δ2000
δ2000
−30◦ 00′ 00′′ −30◦ 00′ 00′′ 0.8

0.25
30′′ 30′′ 0.0
FUV+24µ FUV+24µ
0.00
12s 08s 04s 13h 37m 0013
s h
36m 56s 12s 08s 04s 13h 37m 0013
s h
36m 56s
α2000 α2000
Figure 2.23: ΣSFR in units of M yr−1 kpc−2 (calculated with Eq. 2.4), for the FUV contribution
only (top), the 24µm (middle) and the combination of both (bottom). Right side panels have a
bright foreground point source subtracted, to de-saturate the image.
map is presented in Figure 2.23 (left panels), where we show the ΣSFR with the FUV
contribution only, with the 24µm, and the combination of both.
In Figure 2.23 we noted a very bright and point source at 24µm that was satu-
rating the image, probably a source in the foreground, not part of the galaxy. We
masked this source in the maps of the right panels, and could appreciate much better
the spatial distribution on the ΣSFR across the entire region mapped. The ΣSFR is
mainly dominated by the 24µm emission.
75
Chapter 3
Ram-Pressure stripping
Contents
3.1 Virgo cluster . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.2 Observations . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.3 Data Reduction and spectral line fitting . . . . . . . . . . 80
3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5 Star formation efficiency . . . . . . . . . . . . . . . . . . . 85
3.5.1 HI photometry . . . . . . . . . . . . . . . . . . . . . . . . 87
3.5.2 Hα data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.5.3 K-S relation . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.6 Publication . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
For a long time we have known that galaxies are not isolated systems. On the
contrary, they are highly influenced by their surrounding medium. This can be
directly seen in our local universe, where galaxies that belong to dense clusters have
different properties than the ones in a more isolated medium, even when they have
the same size, mass or age. In particular, one of the most interesting differences
relies on their gas content, since there has been long observational evidence that
spiral galaxies in clusters have less neutral atomic hydrogen than galaxies of the
same morphological type in the field (Haynes & Giovanelli 1984).
No doubt this HI deficiency is strongly related to the fact that those galaxies
belong to a dense galaxy cluster, where galaxies are ruled by the cluster’s dynamics,
and can easily interact with each other. In fact, in overdense cluster environments,
galaxies are significantly transformed, through tidal interactions with other galax-
ies or with the cluster as a whole (e.g. Merritt 1984, Tonnesen et al. 2007), and
interactions with the intra-cluster medium (ICM), which strips them from their gas
content. This effect is most likely to be the result of ram-pressure, i.e the force that
any body, in this case a galaxy, feels when moving through a fluid, like the ICM.
This ram-pressure stripping (RPS) process has been described by Gunn & Gott
(1972), who discussed what might happen if there would be any intergalactic gas
76
left after a cluster had collapsed. The interstellar material in a galaxy would feel
the ram-pressure of the ICM as it moves through the cluster. If this ram-pressure
force is larger than the restoring gravitational force of the disk, the gas of the outer
disk of the galaxy would be stripped off. Even more, they estimate that for a galaxy
moving at a typical velocity of ∼1700 km/s in the Coma cluster, the ISM would be
stripped in one pass. This would explain why so few normal spiral galaxies are seen
in nearby clusters, and so many gas poor, non star forming disk galaxies (Spitzer &
Baade 1951).
The RPS has also been simulated by many groups (Quilis et al. 2000, Vollmer
et al. 2001, Roediger & Hensler 2005; Jáchym et al. 2007), showing the feasibility of
gas stripping due to ram-pressure, and evidence of stripping has been observed in
many cases (Kenney et al. 2004; Chung et al. 2007; Sun et al. 2007, Vollmer et al.
2008) in several galaxies of the nearby clusters Virgo and Abell 3627.
RPS and/or tidal interactions can disperse the interstellar gas (ISM) of galaxies
at large distance, up to 100kpc scales, as shown by the spectacular tail of ionized gas
in Virgo (Kenney et al. 2008), but what is the fate of the stripped gas? According
to the time-scale of the ejection, the relative velocity of the ICM-ISM interaction,
and the environment, it could be first seen as neutral atomic gas (Chung et al. 2009,
Scott et al. 2012, Serra et al. 2013), then ionized gas detected in Hα (Gavazzi et al.
2001, Cortese et al. 2007, Yagi et al. 2007, Zhang et al. 2013), and is finally heated to
X-ray gas temperatures (e.g. Machacek et al. 2005, Sun et al. 2010). In rarer cases,
it can be seen as dense and cold molecular gas, detected as carbon monoxyde (CO)
emission (Vollmer et al. 2005, Dasyra et al. 2012, Jáchym et al. 2014). The presence
of these dense molecular clumps might appear surprising, since the RPS should not
be able to drag them out of their galaxy disks (Nulsen 1982; Kenney & Young 1989).
However they could reform quickly enough in the tail. The survival of these clouds in
the hostile ICM environment, with temperature 107 K and destructive X-rays (e.g.,
Machacek et al. 2004; Fabian et al. 2006; Tamura et al. 2009) is a puzzle, unless
they are self-shielded (e.g. Dasyra et al. 2012, Jáchym et al. 2014). The presence
of cold molecular gas is also observed in rich galaxy clusters, with cool cores. Here
also a multi-phase gas has been detected, in CO, Hα , X-rays and also the strongest
atomic cooling lines (Edge et al. 2010). Ionized gas, together with warm atomic and
molecular gas and cold molecular gas clouds coexist in spatially resolved filaments
around the brightest cluster galaxy, such as in the spectacular prototype Perseus A
(Conselice et al. 2001; Salomé et al. 2006, 2011; Lim et al. 2012).
The survival of molecular clouds was also observed by Braine et al. (2000) in
several tidal tails, and in particular in the interacting system Arp 105 (dubbed the
Guitar), embedded in the X-ray emitting medium of the Abell 1185 cluster (Mahdavi
et al. 1996). Again, the formation in situ of the molecular clouds is favored (Braine
et al. 2000). In the Stephan’s Quintet compact group, where X-ray gas and star
formation have been observed in between galaxies (O’Sullivan et al. 2009), the shock
has been so violent (1000 km/s) that H2 molecules are formed and provide the best
cooling agent, through mid-infrared radiation (Cluver et al. 2010). In this shock,
multi-phases of gas coexist, from cold dense molecular gas to X-ray gas.
77
Does this gas form stars? In usual conditions, inside galaxy disks, the star for-
mation is observed to depend essentially on the amount of molecular gas present
(e.g. Bigiel et al. 2008; Leroy et al. 2013). A Kennicutt-Schmidt (K-S) relation is
observed, roughly linear, between the surface densities of star formation and molec-
ular gas, leading to a depletion time-scale of 2 Gyr. But this relation does not apply
in special regions or circumstances, such as galaxy centers (Casasola et al. 2015),
outer parts of galaxies and extended UV disks (Dessauges-Zavadsky et al. 2014), or
low surface brightness galaxies (Boissier et al. 2008). Little is known on star forma-
tion in gas clouds stripped from galaxies in rich clusters. Boissier et al. (2012) have
put constraints on this process, concluding to a very low star formation efficiency,
lower by an order of magnitude than what is usual in galaxy disks, and even lower
than outer parts of galaxies or in low surface brightness galaxies. It is interesting
to better constrain this efficiency, given the large amount of intracluster light (ICL)
observed today (e.g. Feldmeier et al. 2002, Mihos et al. 2005). These stars could
come from tidal stripping of old stars formed in galaxy disks, or also a large fraction
could have formed in situ, from ram-pressure stripped gas. More intracluster star
formation could have formed in the past (DeMaio et al. 2015). The origin of the
ICL could bring insight on the relative role of galaxy interactions during the cluster
formation, or cluster processing after relaxation.
3.1 Virgo cluster

One of the best laboratories to study the interaction between galaxies is the
Virgo Cluster. Located at ∼18 Mpc and with a mass of 1.2×1015 M (Fouqué
et al. 2001), it hosts more than a thousand spiral and elliptical galaxies, such as
M87, M86 and M49. Many of its galaxies interact with each other in merging or
collissional processes, or interact with the ICM, which strips them from their gas
content. Such is the case in particular of the group of galaxies close to M86, which
shows an spectacular bridge of ionized gas connecting M86 with NGC4438, and a
kpc scale plume of atomic gas being stripped from NGC4388 (Figure 3.1).
This tail northeast of NGC4388 is one of the best ram-pressure tail to probe
molecular gas survival and the star formation efficiency, where X-ray gas has been
mapped (Iwasawa et al. 2003) and young stars have been found (Yagi et al. 2013). It
is located at about 400 kpc in projection from the cluster center M87. NGC 4388 is
moving at a relative velocity redshifted by 1500km/s with respect to M87, and more
than 2800 km/s with respect to the M86 group. This strong velocity may explain
the violent RPS, the high HI deficiency of NGC 4388 (Cayatte et al. 1990) and the
large (∼ 35kpc) emission-line region found by Yoshida et al. (2002), northeast of
the galaxy. The ionized gas has a mass of 105 M , and is partly excited by the
ionizing radiation of the Seyfert 2 nucleus in NGC 4388. The RPS plume is even
more extended in HI (Oosterloo & van Gorkom 2005), up to 110 kpc, with a mass of
3.4 x 108 M . Gu et al. (2013) have found neutral gas in absorption in X-ray, with
column densities 2-3×1020 cm−2 , revealing that the RPS tail is in front of M86. The
high ratio between hot and cold gas in the clouds means that significant evaporation
78
Figure 3.1: Hα + NII image of the Virgo cluster from Kenney et al. (2008). In red: low velocity
emission <500 km/s. In green: high velocity emission >2000 km/s.
79
has proceeded. Yagi et al. (2013) find star-forming regions in the plume at 35 and
66 kpc from NGC 4388, with solar metallicity and age 6 Myrs. Since these stars are
younger than the RPS event, they must have formed in situ.
The aim of the work presented in this chapter is to probe the presence of molecular
gas in the stripped tail northeast of NGC4388, and study its star formation efficiency.
Section 3.2 presents the IRAM-30m CO observations, with their corresponding data
reduction and line fitting in Section 3.3. Results are presented in Section 3.4 and
the gas star formation efficiency discussed in Section 3.5. Finally, the publication of
this work submitted to A&A is presented in Section 3.6.
3.2 Observations
Observation in the HI stream connecting M86 and NGC4388 were done in two
runs. The first one as part of project 195-13 (with 28 hrs of observations) took place
during December 5-8th 2013 in excellent weather conditions, with a τ <0.1 and a
pwv between 0.1 and 3 mm. The second run was part of project 075-14, with 47 hrs
of observations during June 25-30th 2014. The weather conditions in this run were
poor to average, with a τ between 0.2 and 0.6, and a pwv between 3 and 10 mm.
The targets for the first run were chosen in the regions along the HI stream for
having a match of HI, Hα and 250µm emission, to trace atomic gas, ionized gas and
dust respectively. With this criteria six sources were selected and observed (Figure
3.2).
All these targets are listed in Table 3.1, including their coordinates, heliocentric
velocities, and the integration time spent on them.
These observations resulted in two ∼6σ detections of CO(2–1) , in Source-1 and
HaR-2 (Figure 3.3 bottom).
HaR-2 is a source of particular interest since it is so far away from NGC4388 (∼70
kpc), were molecular gas in unexpected, and it has also shown a strong detection in
Hα with the Subaru Telescope (Yagi et al. 2013), therefore this source was chosen
as a central target for the 2014 observations, selecting five more targets close to it,
and where the HI emission seemed brighter (see upper right box in Fig 3.2).
3.3 Data Reduction and spectral line fitting

The data were reduced using the CLASS software from the GILDAS package.
First, a careful inspection of all scans was done, to remove the bad ones. The
approved scans of the same source, CO line and backend were averaged with a
normal time weighting, to obtain one spectra. Then, each spectra was inspected
individually, and in both of its polarizations, to identify a possible CO emission
line. If a detection was found in the spectra of both backends, the best spectra
was chosen (in terms of spectral resolution and S/N) as the final one. The selected
80
Figure 3.2: Targets observed in the Virgo Cluster. Orange: HI contours levels at 1, 5, 10 and
50 ×1019 cm−2 from Oosterloo & van Gorkom (2005). Purple: Hα contours levels at 5, 11 and 50
e− /sec (Kenney et al. 2008). Cyan: 250µm contour levels at 0.01 and 0.1 Jy/beam (Davies et al.
2012). In zoomed regions: targets selected close to HaR-2 for run 075-14.
81
Table 3.1. Targets and Observations.
Source RA(J2000) DEC(J2000) hel. vel. ON+OFF
[hr :m :s ] [deg :m :s ] [km/s] [hrs]
Source-1 12:26:01.3 12:42:30.1 +2500 4.2

Source-2 12:26:04.9 12:45:16.7 +2500 2.9
Source-3 12:26:25.4 12:50:53.6 +2200 3.1
Source-5 12:26:24.5 12:58:14.6 +2000 3.9
HaR-2 12:26:13.7 12:51:36.9 +2230 7.2
HaR-3-4 12:26:13.5 12:43:03.7 +2500 3.8
HaR-2-1 12:26:16.9 12:53:55.6 +2230 4.5
HaR-2-2 12:26:16.9 12:53:19.3 +2230 5.6
HaR-2-3 12:26:16.9 12:52:43.1 +2230 6.2
HaR-2-4 12:26:15.5 12:52:01.1 +2230 8.0
HaR-1 12:26:14.6 12:51:46.4 +2230 3.8
NGC4388 12:25:46.6 12:39:44.0 +2550 1.2
Note. — Sources in the first 6 rows correspond to run 195-13, and

last 6 rows to run 075-14 (although source HaR-2 was observed in both
runs). NGC4388 was observed for calibration purposes. Heliocentric
velocities are referencial, taken from Oosterloo & van Gorkom (2005).
spectra with CO emission are presented in Figures 3.3 and 3.4, which contain both
polarizations, horizontal and vertical, combined.
Baselines were subtracted with polynomials of order 0 and 1 depending on the
source, and antenna temperatures were corrected by the telescope beam efficiency
(Beff ) and forward efficiency (Feff ) to obtain main beam temperatures:
Feff ∗
Tmb = T (3.1)
Beff a
with Beff and Feff equal to 0.78 and 0.94 respectively for 115 GHz, and to 0.59
and 0.92 for 230 GHz.1
Spectra were smoothed with the hanning method to degrade the velocity resolu-
tion until obtain a value no greater than 1/3 of the FWHM line.
Finally, a simple gaussian line was fitted to the line candidate. The CLASS fit
return the velocity position of the line, its FWHM, the peak temperature and the
integrated line intensity.
Such spectra and fitting results for sources with CO detection are presented in
Figures 3.3 and 3.4, and then in Table 3.3.
1
http://www.iram.es/IRAMES/mainWiki/Iram30mEfficiencies
82
Figure 3.3: Top: CO(1–0) and CO(2–1) spectra for the center of NGC4388, observed for calibra-
tion purposes. Bottom: CO(2–1) spectra for HaR-2 and Source-1, taken in the first observing run
195-13. The temperature scale corresponds to main beam temperature (Tmb ).
83
Figure 3.4: Final CO spectra for HaR-2, HaR-1 and HaR-2-4 from run 075-14. Only HaR-2 in
CO(2–1) has data combined from both runs. The temperature scale corresponds to Tmb .
84
Table 3.2. CO(1–0) upper limits at 3σ for sources with no detection.
Source int. time rms ICO MH2 ΣH2
[min] [mK] [K km/s] [106 M ] [M pc−2 ]
Source-2 173 1.2 <0.11 <1.21 <0.48

Source-3 186 1.0 <0.09 <0.99 <0.40
Source-5 231 1.2 <0.11 <1.21 <0.48
HaR-3-4 229 1.0 <0.09 < 0.99 <0.40
HaR-2-1 269 1.1 <0.10 <1.10 < 0.44
HaR-2-2 334 1.1 <0.10 <1.10 <0.44
HaR-2-3 374 0.6 <0.05 <0.55 <0.22
Note. — We assumed a ∆v of 30 km/s to calculate the ICO .
3.4 Results
As a result of this second run, we could confirm the CO detection in HaR-2, and
obtain two more, in HaR-1 and HaR-2-4 (Figure 3.4). Given the large distances of
these sources to NGC4388 (∼70 kpc), it is not likely that the molecular gas was
stripped from the galaxy, and must have formed in situ in the HI gas plume. Details
of these detections and their fitting results are presented in Table 3.3. The range of
linewidths found for these detections is quite large, between 10 and 200 km/s, and
their main beam temperatures were found to be between 1 and 10 mK. H2 masses
were derived from the CO(1–0) line intensity folwing Eq. 1.4, with a Galactic CO
conversion factor of 2.0 × 1020 cm−2 (K km/s)−1 and a correction factor of 1.36 to
account for heavy elements. For the 3 sources with CO(1–0) detected, these masses
were found to be between 0.6 and 3 ×106 M
For the rest of the sources, with no visible CO detections, 3σ upper limits for ICO
where calculated from their rms values, assuming a ∆v = 30 km/s . Such limits are
presented in Table 3.2.
3.5 Star formation efficiency

To estimate how fast is the gas being transformed into stars, we compare the SFR
surface density versus the gas surface density in a K-S relation, to understand the
efficiency of this star formation process. Since these are low gas density regions, we
can expect a large amount of gas in atomic phase, larger than in molecular phase.
Therefore both components, atomic and molecular, need to be taken into account
when estimating a total amount of gas to be converted into stars.
Molecular gas can be directly estimated from the CO(1-0) line intensity, obtained
from our observations. If we consider Eq. 1.4 and average the HI mass in the source’s
area, i.e MHI /πΩ2 , we obtain the H2 surface density:
85
Table 3.3. CO detections in the Virgo Cluster.
Source line backend int. time v0 FWHM resol. Tmb rms ICO MH2
[min] [km/s] [km/s] [km/s] [mK] [mK] [K km/s] [106 M ]
HaR-2 CO(2–1) FTS 268 2224.9±1.2 16.2 ±2.9 4 9.9 2.0 0.17± 0.03 ···
Source-1(1) CO(2–1) WILMA 232 2579.2±2.3 28.7±6.1 5.2 7.0 2.5 0.21±0.04 ···
Source-1(2) CO(2–1) WILMA 232 2731.3±4.0 25.1±12.4 5.2 4.3 2.5 0.12±0.04 ···
HaR-2 CO(1–0) WILMA 318 2200.3±8.6 72.6±15.9 10.4 3.5 2.1 0.27±0.06 3.24
··· CO(2–1) WILMA 586 2224.3±2.7 26.6±7.2 10.5 5.0 1.8 0.14±0.03 ···
HaR-1 CO(1–0) WILMA 229 2212.2±4.5 32.8±13.8 10.4 5.0 2.3 0.18±0.05 2.11
··· CO(2–1) WILMA 229 2223.2±7.8 80.9±15.15 10.4 6.0 3.2 0.52± 0.09 ···
HaR-2-4 CO(1–0) WILMA 665 2402.6±2.3 10.4±62.1 10.4 4.2 1.5 0.47±0.02 0.57
86
··· CO(2–1) WILMA 665 2281.7±26.4 200.4±45.3 20.8 1.4 2.8 0.29±0.07 ···
NGC4388 CO(1–0) WILMA 87 2501.7±2.7 278.6±5.9 5.2 52.2 5.7 15.47±0.30 186
··· CO(2–1) WILMA 82 2492.0±3.5 273.3±7.4 5.2 67 9.4 19.49±0.49 ···
H2 masses were directly calculated from the CO(1–0) emission using Eq.1.4, using a source radius R equivalent to the CO(1–0) beamsize
radius at the distance for the source (17.5 Mpc from Mei et al. 2007).
Note. — First three rows correspond to observing run 195-13. Both components shown in Figure 3.3 for Source-1 are listed. H2 masses
were calculated as MH2 [M ]=4.4πR2 [pc] ICO [K km/s], which uses a Galactic CO conversion factor of 2.0×1020 [cm−2 (K km/s)−1 ] and
a correction factor of 1.36 to account for He. A distance to the Virgo Cluster of 17.5 Mpc (Mei et al. 2007) was used to calculate the
CO(1–0) beam radius as source radius. HaR-2 includes data taken in the first run only for CO(1–0), and data combined from both runs
for CO(2–1).
ΣH2 [M pc−2 ] = 4.4ICO(1−0) [K km s−1 ] (3.2)
These values are listed in column 2 of Table 3.4 for the sources with CO(1–0)
detections, including upper limits for the sources with no CO detection using Table
3.2.
3.5.1 HI photometry
The mass of atomic gas can be directly estimated from the HI column density
(NHI ), which integrated inside the source’s area gives the amount of HI atoms present
in that area. Thus, by multiplying by the hydrogen mass:
Z
MHI = µmH NH dA (3.3)
where µ = 1.36 is the correction factor to account for heavy elements and mH is
the mass of one hydrogen atom (8.4×10−58 M ). Since the area differential is related
with the solid angle differential as dΩ=dA/D2 , then
Z
2
MHI = µmH D NH dΩ (3.4)
where the solid angle Ω used here corresponds to the CO(1–0) beamsize radius
of 1100 . To measure the HI mass for the sources in our sample, we used the HI
column density map of Oosterloo & van Gorkom (2005) for the plume northeast of
NGC4388 (Figure 3.5). The corresponding integrated column densities are listed in
column 3 of Table 3.4.
Finally, the HI surface density is calculated by averaging the HI mass in the
surface area to be considered, i.e ΣHI = MHI /Ω, where Ω is the CO(1–0) beam solid
angle at the distance of the source (i.e Ω=2.74 ×106 pc2 at D=17.5 Mpc). This
surface densities are presented in column 4 of Table 3.4.
3.5.2 Hα data
The Hα emission seen in the plume northeast of NGC4388 shows evidence of star
forming regions (HII regions). Therefore it can be used to estimate the rate at which
the gas is being converted into stars. The SFR can be calculated directly from the
Hα luminosity by following Kennicutt & Evans (2012):
87
Source-5
HaR-2
Source-3
Source-2
HaR-3-4
Source-1
NGC4388
Figure 3.5: HI column density map from Oosterloo & van Gorkom (2005) taken with the West-
erbork radiotelescope. Our sources are marked in the same way as in Figure 3.2. Aperture
photometry was done for all sources inside apertures of 2200 diameter, which here look as ovals
since the image has rectangular pixels.
88
Table 3.4. K-S relation values.
R
Source Σ H2 NHI dΩ ΣHI log(LHα ) log(ΣSFR ) τdep (H2 )
[M pc−2 ] [10 cm−2 ]
12 [M pc−2 ] [erg s−1 ] [M yr−1 kpc−2 ] [yr]
Source-1 0.93 0.33 0.40 <35.51 < -6.19 >1.4×1012

Source-2 <0.48 0.41 0.50 <35.51 < -6.19 ···
Source-3 <0.40 0.56 0.68 <35.51 < -6.19 ···
Source-5 <0.48 0.60 0.69 37.51 -4.20 <7.7×109
HaR-2 1.18 2.12 2.5 37.10 -4.61 4.8×1010
HaR-3-4 <0.40 ··· ··· 37.75 -3.96 <3.6×109
HaR-2-1 <0.44 5.56 6.77 <35.51 < -6.19 ···
HaR-2-2 <0.44 5.13 6.24 <35.51 < -6.19 ···
HaR-2-3 <0.22 3.67 4.46 <35.51 < -6.19 ···
HaR-2-4 2.09 3.08 3.74 < 35.51 < -6.19 >3.2×1012
HaR-1 0.77 2.05 2.49 35.89 -5.82 5.1×1011
Note. — Surface densities consider a solid angle Ω=2.74 kpc2 , equivalent to the CO(1–0) beam-
size at the distance of the source (17.5 Mpc fom Mei et al. 2007). Depletion times are calculated as
τdep = ΣH2 /ΣSFR .
log SFR = log LHα [erg s−1 ] − 41.27 (3.5)
which uses the initial mass function (IMF) of Kroupa & Weidner (2003). Once
again, if we average this value over the surface area to be taken into account, we
obtain the SFR surface density: ΣSFR = SFR/Ω, with Ω the CO(1–0) beam solid
angle.
Hα luminosities for HaR-1, HaR-2 and HaR-3-4 were obtained from Yagi et al.
(2013) with the Subaru Prime Focus Camera. Such values are presented in column
5 of Table 3.4 with their corresponding ΣSFR in column 6 following Eq. 3.5
For the remaining sources in our sample we used the Hα+[NII] map of the Virgo
cluster from Kenney et al. (2008), presented in Figure 3.6. We used the Hα lumi-
nosities of HaR-2, HaR-3 and HaR-4 of Yagi et al. (2013) to calibrate the Hα+[NII]
emission to Hα intensities, assuming that all HII regions have the same NII/Hα
ratio.
With this calibration, and using an aperture photometry of 100 diameter (similar
to the seeing of the observations of Yagi et al. 2013) we calculated an LHα for Source-5
(Table 3.4). For the rest of the sources with no visible emission, we calculated upper
limits from the noise level of the map, estimated in 4.8×10−7 erg s−1 cm−2 sr−1 . Inside
a 100 aperture and at the distance of 17.5 Mpc, this limit translates to 3.24×1035 erg
s−1 . Then by using Eq. 3.5 and the averaging the SFR in the CO(1–0) beam solid
angle Ω, we obtain an upper limit of 6.49×10−7 M yr−1 kpc−2 .
All these values are presented in Table 3.4
89
Source-5
M86
HaR-2
HaR-2
Source-3
Source-2
HaR-3-4
Source-1
Figure 3.6: Hα+[NII] map of the Virgo cluster from Kenney et al. (2008) taken with the KPNO
telescope. Our sources are marked in the same way as in Figure 3.2. Circles in this image are
2200 width, but the aperture photometry was done in a 100 aperture, to be consistent with the Hα
data from Yagi et al. (2013). This map was calibrated with Yagi’s data and used to calculate the
Lα of Source-5, as well as the upper limits for Source-1, Source-2, Source-3, HaR-2-1, HaR-2-2,
HaR-2-3 and HaR-2-4 from the map’s noise level.
90
Figure 3.7: K-S relation for sources in Table 3.4, with filled circles and arrows for values and
upper limits respectively. Red markers consider only H2 for Σgas , black ones only HI, and blue
ones the sum of both. Figure adapted from Jáchym et al. (2014), where their sources ESO137-001,
001-A, 001-B and 001-C have been plotted in a similar way than ours: red circles for H2 gas,
black left arrows for HI upper limits, and blue ”error bars” to account for both gas components.
Colored contours account for the spiral galaxies from Bigiel et al. (2008) (green, orange, red and
purple for 1, 2, 5 and 10 sampling points per 0.05 dex respectively). Black markers with error
bars correspond to the running medians in ΣSF R as a function of σH2 of 30 nearby galaxies from
Bigiel et al. (2011). Shaded ovals represent the data from the outer parts of XUV disk galaxies:
NGC4625 and NGC6946 from Watson et al. (2014) (priv. communication), and M63 (NGC5055)
from Dessauges-Zavadsky et al. (2014), taking only H2 into account in all of them. The dashed
vertical line shows the 9 M pc−2 threshold at which the atomic gas saturates. Dashed inclined
lines represent “isochrones” of constant star formation efficiencies, indicating the depletion times
τdep of 108 , 109 and 1010 years to consume all the gas. The dashed red isochrone marks a depletion
time equal to the age of the universe, as one Hubble time (13.8 Gyr). A representative shift of the
HaR-1 marker for ΣH2 is drawn, to show the ”effective” molecular gas density at which stars would
formed in this region if we consider a beam correction factor of +2.68 in log space, to convert our
2200 beam to a ∼100 beam as in the Hα data.
91
3.5.3 K-S relation
The SFR surface densities are plotted as a function of the gas surface densities
to construct a K-S relation in Figure 3.7, using the values from Table 3.4. We
have plotted separately, the atomic and molecular gas component of Σgas , along
with the combination of both, in red, black and blue markers respectively. Arrows
denote the upper limits values from Table 3.4. In this Figure, adapted from Jáchym
et al. (2014), we can compare our sources with with theirs in the Norma Cluster,
as well as with the sample of spiral galaxies from Bigiel et al. (2008) (in coloured
contours), and the sample of 30 nearby galaxies from Bigiel et al. (2011), plotted as
the running medians of ΣSFR as a function of ΣH2 , and with a typical depletion time
(τdep ) of ∼2.3 Gyr. Additionally, shaded regions have been included to represent
regions from the outskirts of XUV disk galaxies. NGC4625 and NGC6946 data was
taken from Watson et al. (2014) (priv. communication), including IRAM-30m CO
observations, and Hα luminosities measured within a 600 aperture. M63 (NGC5055)
data corresponds to the bright UV region located at 1.35r25 in Dessauges-Zavadsky
et al. (2014), with IRAM 30-m CO data, and a SFR measured from the FUV and
24µm emission.
“Isochrones” of constant star formation efficiencies are also shown to indicate
the depletion times of 108 , 109 and 1010 years to consume all the gas, including an
additional red isochrone to mark the age of the universe as one Hubble time (i.e
τdep = 13.8Gyr).
Contrary to the photometry done in the HI data, the Hα data was not measured
in a 2200 diameter aperture, as the CO(1–0) FWHM, but in an aperture of 100 in
diameter, similar to the seeing of those observations. Since we are averaging this
Hα emission in a 2200 aperture to calculate the ΣSFR , we could be diluting the real
surface density of the gas being converted into stars. To correct for this beam
dilution, we have shifted in a representative way one of the points in Figure 3.7
to a fictitious ΣH2 , corresponding to a source’s solid angle of 100 in diameter. This
correction translates in a +2.68 shift in log space, and is a representation of the real
gas surface density at which stars would be formed, but always with the same τdep .
From Figure 3.7 we can see that our sources have extremely low SFRs in com-
parison with the nearby spiral galaxies, and are only comparable with the most
outer clumps in the Hα /X-ray tail of the ISM stripped galaxy ESO137-001 in the
Norma Cluster (Jáchym et al. 2014) and the XUV disk galaxies from Watson et al.
(2014) and Dessauges-Zavadsky et al. (2014). We calculated the corresponding de-
pletion times, i.e the amount of time needed to consume all the gas available to be
converted into stars (τdep = Σgas /ΣSFR ), obtaining values significantly large. For
example, HaR-1 and HaR-2 have τdep values of 2.1×1012 and 1.5×1011 years respec-
tively, to consume all the amount of gas present (HI+H2 ). These values transform
into 1.6×1012 and 1.0×1011 years if we consider only the atomic gas component, and
into 5.1×1011 and 4.8×1010 years if we consider only the molecular one. Such values
are quite large in comparison with the typical τdep of ∼2 Gyrs for spiral galaxies,
and are even larger than a Hubble time by up to 2 orders of magnitude. In Table
3.4, depletion times of H2 that can be calculated are listed in column 7. These high
92
values of τdep , much larger than the age of the universe, indicate that this molecular
gas will no be converted into stars, but remain in a gaseous phase and later join the
ICM.
The extremely low SF efficiency of our sources seems to fall off the linearity of
the K-S relation for typical spiral galaxies at higher gas densities, a result previ-
ously reported in other low gas density environments, such as XUV disk galaxies
(Dessauges-Zavadsky et al. 2014). Watson et al. (2014) presents a different conclu-
sion for their results in XUVs, with a typical SFR in agreement with the K-S linear
regime, but they take into account the 24µm emission in the SFR, and neglect the
contribution of heavy elements in the ΣH2 . We see that when we correct by these
differences to make their data analytically compatible with ours (i.e neglect the
24µm emission and correct for heavy elements), their data points in the K-S plot
are comparable to ours.
Therefore, our results help to show that the linearity between the SFR and the
gas surface density, so clear at high gas surface densities (> 9 M pc−2 ) for normal
spiral galaxies, cannot be extrapolated to lower densities, below the HI-H2 threshold,
where the star formation is extremely inefficient, and the molecular gas, even though
present, does not necessarily form stars .
3.6 Publication
The corresponding publication of this work was submitted to A&A in May 2015
and is presented here.
93
Astronomy & Astrophysics manuscript no. virgo_ram_pressure c
ESO 2015
May 17, 2015
Ram Pressure stripping in the Virgo Cluster

C. Verdugo1 , F. Combes1, 2 , K. Dasyra1, 3 , P. Salomé1 and J. Braine4
1
Observatoire de Paris, LERMA (CNRS:UMR8112), 61 Av. de l’Observatoire, F-75014, Paris, France e-mail:
[email protected]
2
Collège de France, 11 Place Marcelin Berthelot, 75005 Paris, France
3
Department of Astrophysics, Astronomy & Mechanics, Faculty of Physics, University of Athens, Panepistimiopolis Zografos
15784, Greece
4
Univ. Bordeaux, Laboratoire d’Astrophysique de Bordeaux, (CNRS:UMR5804) 33270 Floirac, France
Received 2015; accepted 2015
ABSTRACT
Gas can be violently stripped from their galaxy disks in rich clusters, and be dispersed over 100kpc-scale tails or plumes. Young stars
have been observed in these tails, suggesting they are formed in situ. This will contribute to the intracluster light, in addition to tidal
stripping of old stars. We want to quantify the efficiency of intracluster star formation. We present CO(1–0) and CO(2–1) observations,
made with the IRAM-30m telescope, towards the ram-pressure stripped tail northeast of NGC4388 in Virgo. HII regions found all
along the tails, together with dust patches have been targeted. We detect molecular gas in 4 positions along the tail, with masses
between 6x105 to 3x106 M . Given the large distance from the NGC 4388 galaxy, the molecular clouds must have formed in situ,
from the HI gas plume. We compute the relation between surface densities of star formation and molecular gas in these regions, and
find that the star formation has very low efficiency. The corresponding depletion time of the molecular gas can be up to 500 Gyr and
more. Since this value exceeds a by far Hubble time, this gas will not be converted into stars, and will stay in a gaseous phase to join
the intracluster medium.
Key words. Galaxies: evolution — Galaxies: clusters: Individual: Virgo — Galaxies: clusters: intracluster medium — Galaxies:
interactions — Galaxies: ISM
1. Introduction they could reform quickly enough in the tail. The survival of
these clouds in the hostile ICM environment, with temperature
In overdense cluster environments, galaxies are significantly 107 K and destructive X-rays (e.g., Machacek et al. 2004; Fabian
transformed, through tidal interactions with other galaxies or et al. 2006; Tamura et al. 2009) is a puzzle, unless they are self-
with the cluster as a whole (e.g. Merritt 1984, Tonnesen et al. shielded (e.g. Dasyra et al. 2012, Jáchym et al. 2014). The pres-
2007), and interactions with the intra-cluster medium (ICM), ence of cold molecular gas is also observed in rich galaxy clus-
which strips them from their gas content. This ram-pressure ters, with cool cores. Here also a multi-phase gas has been de-
stripping (RPS) process has been described by Gunn & Gott tected, in CO, Hα, X-rays and also the strongest atomic cooling
(1972) and simulated by many groups (Quilis et al. 2000, lines (Edge et al. 2010). Ionized gas, together with warm atomic
Vollmer et al. 2001, Roediger & Hensler 2005; Jáchym et al. and molecular gas and cold molecular gas clouds coexist in spa-
2007). Evidence of stripping has been observed in many cases tially resolved filaments around the brightest cluster galaxy, such
(Kenney et al. 2004; Chung et al. 2007; Sun et al. 2007, Vollmer as in the spectacular prototype Perseus A (Conselice et al. 2001;
et al. 2008). RPS and/or tidal interactions can disperse the in- Salomé et al. 2006, 2011; Lim et al. 2012).
terstellar gas (ISM) of galaxies at large distance, up to 100kpc
scales, as shown by the spectacular tail of ionized gas in Virgo The survival of molecular clouds was also observed by
(Kenney et al. 2008). Braine et al. (2000) in several tidal tails, and in particular in the
interacting system Arp 105 (dubbed the Guitar), embedded in
What is the fate of the stripped gas? According to the time-
the X-ray emitting medium of the Abell 1185 cluster (Mahdavi
scale of the ejection, the relative velocity of the ICM-ISM in-
et al. 1996). Again, the formation in situ of the molecular clouds
teraction, and the environment, it could be first seen as neutral
is favored (Braine et al. 2000). In the Stephan’s Quintet compact
atomic gas (Chung et al. 2009, Scott et al. 2012, Serra et al.
group, where X-ray gas and star formation have been observed in
2013), then ionized gas detected in Hα (Gavazzi et al. 2001,
between galaxies (O’Sullivan et al. 2009), the shock has been so
Cortese et al. 2007, Yagi et al. 2007, Zhang et al. 2013), and
violent (1000 km/s) that H2 molecules are formed and provide
is finally heated to X-ray gas temperatures (e.g. Machacek et al.
the best cooling agent, through mid-infrared radiation (Cluver
2005, Sun et al. 2010). In rarer cases, it can be seen as dense and
et al. 2010). In this shock, multi-phases of gas coexist, from cold
cold molecular gas, detected as carbon monoxyde (CO) emission
dense molecular gas to X-ray gas.
(Vollmer et al. 2005, Dasyra et al. 2012, Jáchym et al. 2014). The
presence of these dense molecular clumps might appear surpris- Does this gas form stars? In usual conditions, inside galaxy
ing, since the RPS should not be able to drag them out of their disks, the star formation is observed to depend essentially on
galaxy disks (Nulsen 1982; Kenney & Young 1989). However the amount of molecular gas present (e.g. Bigiel et al. 2008;
Leroy et al. 2013). A Kennicutt-Schmidt (K-S) relation is ob-
Send offprint requests to: C. Verdugo served, roughly linear, between the surface densities of star for-
Article number, page 1 of 8
A&A proofs: manuscript no. virgo_ram_pressure
mation and molecular gas, leading to a depletion time-scale of 2. Observations and data reduction
2 Gyr. But this relation does not apply in special regions or cir-
cumstances, such as galaxy centers (Casasola et al. 2015), outer CO observations along the HI stream connecting NGC4388 and
parts of galaxies and extended UV disks (Dessauges-Zavadsky M86 were done with the IRAM 30-mt telescope at Pico Veleta,
et al. 2014), or low surface brightness galaxies (Boissier et al. Spain, in two separate runs. The first run was part of the project
2008). Little is known on star formation in gas clouds stripped 195-13, with 28 hours of observation, and took place between the
from galaxies in rich clusters. Boissier et al. (2012) have put 5th and 8th of December 2013, with excellent weather conditions
constraints on this process, concluding to a very low star for- (τ < 0.1 and a pwv between 0.1 and 3mm). The second run
mation efficiency, lower by an order of magnitude than what is was project 075-14, with 47 hours of observations between June
usual in galaxy disks, and even lower than outer parts of galaxies 25th-30th 2014, and had poor to average weather conditions (τ
or in low surface brightness galaxies. It is interesting to better between 0.2 and 0.6 and a pwv between 3 and 10 mm).
constrain this efficiency, given the large amount of intracluster All observations were done with the EMIR receiver in the
light (ICL) observed today (e.g. Feldmeier et al. 2002, Mihos E0/E2 configuration, allowing us to observe simultaneously
et al. 2005). These stars could come from tidal stripping of old CO(1–0) and CO(2–1) at 115.271 and 230.538 GHz respectively.
stars formed in galaxy disks, or also a large fraction could have The telescope half-power beam widths at these frequencies are
formed in situ, from ram-pressure stripped gas. More intraclus- 2200 and 1100 respectively. The observing strategy consisted in
ter star formation could have formed in the past (DeMaio et al. single ON+OFF pointings per each target, with wobbler switch-
2015). The origin of the ICL could bring insight on the relative ing.
role of galaxy interactions during the cluster formation, or clus- Targets along the HI stream were selected for having a match
ter processing after relaxation. of HI, Hα and 250µ emission (Figure 1). With this criteria 6 tar-
gets where selected for the first run 195-13, and are listed in
Table 1 (first 6 rows). As a result of this run, only two sources
showed CO detection: Source-1 and HaR-2. Since HaR-2 is of
1.1. The tail northeast of NGC4388 particular interest for being so far away from both galaxies and
for having a strong Hα detection (Yagi et al. 2013), it was cho-
One of the best ram-pressure tails to probe molecular gas sur- sen as a central target around which other 5 extra targets were
vival under extreme conditions, and the star formation efficiency selected for the second run 075-14 (second half of Table 1), fol-
is the RPS tail north of NGC 4388 in the Virgo cluster south of lowing the path of an HI peak (Figure 1 top right box)
M86, where X-ray gas has been mapped (Iwasawa et al. 2003)
and young stars have been found (Yagi et al. 2013). It is located Table 1. Targets and Observations
at about 400 kpc in projection from the cluster center M87. NGC
4388 is moving at a relative velocity redshifted by 1500km/s Source RA(J2000) DEC(J2000) hel. vel. ON+OFF
with respect to M87, and more than 2800 km/s with respect to the [hr :m : s ] [deg :m : s ] [km/s] [hrs]
M86 group. This strong velocity may explain the violent RPS, Source-1 12:26:01.3 12:42:30.1 +2500 4.2
the high HI deficiency of NGC 4388 (Cayatte et al. 1990) and Source-2 12:26:04.9 12:45:16.7 +2500 2.9
the large (∼ 35kpc) emission-line region found by Yoshida et al. Source-3 12:26:25.4 12:50:53.6 +2200 3.1
(2002), northeast of the galaxy. The ionized gas has a mass of Source-5 12:26:24.5 12:58:14.6 +2000 3.9
105 M , and is partly excited by the ionizing radiation of the HaR-2 12:26:13.7 12:51:36.9 +2230 7.2
Seyfert 2 nucleus in NGC 4388. The RPS plume is even more HaR-3-4 12:26:13.5 12:43:03.7 +2500 3.8
extended in HI (Oosterloo & van Gorkom 2005), up to 110 kpc, HaR-2-1 12:26:16.9 12:53:55.6 +2230 4.5
with a mass of 3.4 x 108 M . Gu et al. (2013) have found neu- HaR-2-2 12:26:16.9 12:53:19.3 +2230 5.6
tral gas in absorption in X-ray, with column densities 2-3×1020 HaR-2-3 12:26:16.9 12:52:43.1 +2230 6.2
cm−2 , revealing that the RPS tail is in front of M86. The high ra- HaR-2-4 12:26:15.5 12:52:01.1 +2230 8.0
tio between hot and cold gas in the clouds means that significant HaR-1 12:26:14.6 12:51:46.4 +2230 3.8
evaporation has proceeded. Yagi et al. (2013) find star-forming NGC4388 12:25:46.6 12:39:44.0 +2550 1.2
regions in the plume at 35 and 66 kpc from NGC 4388, with
solar metallicity and age 6 Myrs. Since these stars are younger Notes. Sources in the first 6 rows correspond to run 195-13, and last
than the RPS event, they must have formed in situ. 6 rows to run 075-14 (although source HaR-2 was observed in both
runs). NGC4388 was observed for calibration purposes. Heliocentric
In the present paper we present CO detections in the ram- velocities are referencial, taken from Oosterloo & van Gorkom (2005).
pressure stripped gas northeast of NGC 4388. In a previous pa-
per, we have already found molecular gas in a ionized gas tail Concerning the spectral resolution of our data, during the
south of M86 (Dasyra et al. 2012), and discussed its survival observations both FTS and WILMA backends were used, simul-
conditions. We here study the link between new stars formed and taneously. The FTS backend has a spectral resolution of 195 kHz
molecular gas, to derive the star formation efficiency. In the RPS and a bandwidth of 32 GHz including both polarizations. At 115
plume, a significant fraction of the Hα emission could originate GHz these values correspond to 0.5 and 83200 km/s, and at 230
from the ionized gas in the outer layers of molecular clouds (Fer- GHz to 0.25 and 41600 km/s. As for the WILMA backend, we
land et al. 2009). This makes the Hα lumps good tracers of star obtained a spectral resolution of 2 MHz and a bandwidth of 16
formation in an RPS tail, to probe the efficiency of the process of GHz. At 115 GHz this translates to 5.2 and 41600 km/s, and to
formation of intracluster stars. Section 2 presents the IRAM-30m 2.6 and 20800 km/s at 230 GHz.
observations, Section 3 the results obtained, which are discussed The data were reduced using the CLASS software from the
in Section 4. GILDAS package. First, a careful inspection of all scans was
done, to remove the bad ones. The approved scans of the same
In the following, we assume a distance of 17.5 Mpc to the source, CO line and backend were averaged with a normal time
Virgo cluster (Mei et al. 2007). weighting, to obtain one spectra. Then, each spectra was in-
C. Verdugo et al.: CO in the Virgo Cluster
Source-5 HaR-2-1
M86
HaR-2-2
56′ HaR-2-3
HaR-2-4
52′ HaR-2 HaR-1

HaR-2
Dec (J2000)
Source-3
48′
Source-2
44′ HaR-3-4
Source-1
NGC4388
+12◦40′
30s 15s 26m00s 12h25m45s

RA (J2000)
Fig. 1. Targets observed in the Virgo Cluster. Orange: HI contours levels at 1, 5, 10 and 50 ×1019 cm−2 from Oosterloo & van Gorkom (2005).
Purple: Hα contours levels at 5, 11 and 50 e− /sec (Kenney et al. 2008). Cyan: 250µm contour levels at 0.01 and 0.1 Jy/beam (Davies et al. 2012).
In zoomed regions: targets selected close to HaR-2 for run 075-14 (red circles). Circles enclosing targets are 2200 width, as the CO(1–0) HPBW.
spected individually, and in both of its polarizations, to identify a run) and then in Table 2. For the rest of the sources, with no
possible CO emission line. If a detection was found in the spec- visible CO detections, 3σ upper limits for ICO where calculated
tra of both backends, the best spectra was chosen (in terms of from their rms values, assuming a ∆v = 30 km/s . Such limits are
spectral resolution and S/N) as the final one. The selected spec- presented in Table 3.
tra with CO emission are presented in Figures 2 and 3, which
contain both polarizations, horizontal and vertical, combined.
Baselines were subtracted with polynomials of order 0 and 3. Results
1 depending on the source, and antenna temperatures were cor-
rected by the telescope beam and forward efficiencies1 to obtain After selecting the final spectra for every source with a visible
main beam temperatures. Spectra were smoothed with the han- detection and fitting their gaussian profiles, line parameters are
ning method to degrade the velocity resolution until obtaining a calculated and presented in Table 2. H2 masses were derived
value no greater than 1/3 of the FWHM line. from the CO(1–0) line intensity, using a Galactic CO conversion
Finally, a simple gaussian line was fitted to the line candi- factor of 2.0 × 1020 [cm−2 (K km/s)−1 ] and a correction factor of
date. The CLASS fit return the velocity position of the line, its 1.36 to account for heavy elements:
FWHM, the peak temperature and the integrated line intensity.
Such spectra and fitting results for sources with CO detection
are presented in Figures 2 (from the first run) and 3 (from second
1
http://www.iram.es/IRAMES/mainWiki/Iram30mEfficiencies MH2 [M ] = 4.4πR2 [pc] ICO(1−0) [K km/s] (1)

Fig. 2. Top: Spectra taken at the center of NGC4388, observed for calibration purposes. Bottom: Spectra from observing run 195-13 that showed
CO emission. Spectra information and line parameters are presented in Table 2. The temperature scale corresponds to main beam temperature.
Table 2. CO detections and their line parameters.
Source line backend int. time v0 FWHM resol. Tmb rms ICO MH2
[min] [km/s] [km/s] [km/s] [mK] [mK] [K km/s] [106 M ]
HaR-2 CO(2–1) FTS 268 2224.9±1.2 16.2 ±2.9 4 9.9 2.0 0.17± 0.03
Source-1(1) CO(2–1) WILMA 232 2579.2±2.3 28.7±6.1 5.2 7.0 2.5 0.21±0.04
Source-1(2) CO(2–1) WILMA 232 2731.3±4.0 25.1±12.4 5.2 4.3 2.5 0.12±0.04
HaR-2 CO(1–0) WILMA 318 2200.3±8.6 72.6±15.9 10.4 3.5 2.1 0.27±0.06 3.24
CO(2–1) WILMA 586 2224.3±2.7 26.6±7.2 10.5 5.0 1.8 0.14±0.03
HaR-1 CO(1–0) WILMA 229 2212.2±4.5 32.8±13.8 10.4 5.0 2.3 0.18±0.05 2.11
CO(2–1) WILMA 229 2223.2±7.8 80.9±15.15 10.4 6.0 3.2 0.52± 0.09
HaR-2-4 CO(1–0) WILMA 665 2402.6±2.3 10.4±62.1 10.4 4.2 1.5 0.47±0.02 0.57
CO(2–1) WILMA 665 2281.7±26.4 200.4±45.3 20.8 1.4 2.8 0.29±0.07
NGC4388 CO(1–0) WILMA 87 2501.7±2.7 278.6±5.9 5.2 52.2 5.7 15.47±0.30 186
CO(2–1) WILMA 82 2492.0±3.5 273.3±7.4 5.2 67 9.4 19.49±0.49
Notes. First three rows correspond to observing run 195-13. Both components shown in Figure 2 for Source-1 are listed. H2 masses were calculated
as MH2 [M ]=4.4πR2 [pc] ICO [K km/s], which uses a Galactic CO conversion factor of 2.0×1020 [cm−2 (K km/s)−1 ] and a correction factor of 1.36
to account for heavy elements. A distance to the Virgo Cluster of 17.5 Mpc (Mei et al. 2007) was used to calculate the CO(1–0) beam radius as
source radius. HaR-2 includes data taken in the first run only for CO(1–0), and data combined from both runs for CO(2–1).
where the source’s radius R corresponds to the CO(1–0) 3.1. Star formation efficiency
beamsize radius at the distance of the source (17.5 Mpc from
Mei et al. 2007). To estimate how fast is the gas being transformed into stars, we
compare the SFR surface density versus the gas surface den-
sity in a K-S relation, to understand the efficiency of this star
formation process. Since these are low gas density regions, we
can expect a large amount of gas in atomic phase, greater than
in molecular phase. Therefore both components, atomic and
Fig. 3. Final CO spectra for HaR-2, HaR-1 and HaR-2-4 from run 075-14. Only HaR-2 in CO(2–1) has data combined from both runs. All spectra
have both polarizations combined. Spectra information and line parameters are presented in Table 2. The temperature scale corresponds to main
beam temperature.
Table 3. CO(1–0) upper limits at 3σ for sources with no detection. molecular, need to be taken into account when estimating a total
amount of gas to be converted into stars.
Source int. time rms ICO a MH2 ΣH2
Molecular gas can be directly estimated from the CO(1-0)
[min] mK [K km/s] [106 M ] [M pc−2 ]
line intensity, obtained from our observations. If we take Eq. 1,
Source-2 173 1.2 <0.11 <1.21 <0.48
Source-3 186 1.0 <0.09 <0.99 <0.40 and we divide it by the source area (πR2 ), we obtain the H2 sur-
Source-5 231 1.2 <0.11 <1.21 <0.48 face density:
HaR-3-4 229 1.0 <0.09 < 0.99 <0.40
HaR-2-1 269 1.1 <0.10 <1.10 < 0.44
HaR-2-2 334 1.1 <0.10 <1.10 <0.44
HaR-2-3 374 0.6 <0.05 <0.55 <0.22
(a)
We assumed a ∆v of 30 km/s. ΣH2 [M pc−2 ] = 4.4ICO(1−0) [K km s−1 ] (2)

These values are listed in column 2 of Table 4 for the sources a 2200 aperture to calculate the ΣSFR , we could be diluting the real
with CO(1–0) detections, including upper limits for the sources surface density of the gas being converted into stars. To correct
with no CO detection using Table 3. for this beam dilution, as a representation we have shifted one
For the atomic gas component, we estimated the amount of of the points in Figure 4 to a fictitious ΣH2 , corresponding to a
HI from the HI column density map of the NGC4388 plume from source’s solid angle of 100 in diameter. This correction translates
Oosterloo & van Gorkom (2005). The HI mass is derived from in a +2.68 shift in log space, and is a representation of the real
the integrated amount of NH inside the source solid angle: gas surface density at which stars would be formed, but always
with the same τdep .
Z Z From Figure 4 we can see that our sources have extremely
low SFRs in comparison with the nearby spiral galaxies, and
MHI = µmH NH dA = µmH D2 NH dΩ (3) are only comparable with the most outer clumps in the Hα /X-
ray tail of the ISM stripped galaxy ESO137-001 in the Norma
Aperture photometry was done in the NH map of Oosterloo Cluster (Jáchym et al. 2014) and the XUV disk galaxies from
& van Gorkom (2005), inside a circle of 1100 radius to obtain the Watson et al. (2014) and Dessauges-Zavadsky et al. (2014). We
integrated column density. These values are listed in column 3 obtain depletion times (τdep = Σgas /ΣSFR ) significantly large.
of Table 4. Then, by dividing Eq. 3 by the CO(1–0) beam solid For example, HaR-1 and HaR-2 have τdep values of 2.1×1012
angle Ω = πR2 , we obtain the HI surface densities ΣHI listed in and 1.5×1011 years respectively, to consume all the amount of
column 4 of Table 4. gas present (HI+H2 ). These values transform into 1.6×1012 and
Finally, we estimate the SFR surface density ΣSFR directly 1.0×1011 years if we consider only the atomic gas component,
from the Hα emission in these regions. From Kennicutt & Evans and into 5.1×1011 and 4.8×1010 years if we consider only the
(2012): molecular one. Such values are quite large in comparison with
the typical τdep of ∼2 Gyrs for spiral galaxies, and are even larger
than a Hubble time by up to 2 orders of magnitude. In Table 4,
log SFR[M yr−1 ] = log LHα [erg s−1 ] − 41.27 (4) depletion times of H2 that can be calculated are listed in column
7.
which, divided by the CO(1–0) beam solid angle Ω, gives The extremely low SF efficiency of our sources seems to fall
the SFR surface density ΣSFR . Hα luminosities for HaR-1, HaR- off the linearity of the K-S relation for typical spiral galaxies at
2 and HaR-3-4 were obtained from Yagi et al. (2013), and are higher gas densities, a result previously reported in other low gas
listed in Table 4, along with the corresponding ΣSFR . For the re- density environments, such as XUV disk galaxies (Dessauges-
maining sources we used the Hα map from Kenney et al. (2008), Zavadsky et al. 2014). Watson et al. (2014) presents a differ-
using an aperture photometry of ∼100 in diameter, similar to the ent conclusion for their results in XUVs, with a typical SFR in
seeing of the Hα observations from Yagi et al. (2013). For the agreement with the K-S linear regime, but they take into account
sources with no visible detection in this map, an upper limit was the 24µm emission in the SFR, and neglect the contribution of
calculated form the noise level of Kenney’s map, estimated in heavy elements in the ΣH2 . We see that when we correct by these
4.8 × 10−7 erg s−1 cm−2 sr−1 . differences to make their data analytically compatible with ours
The SFR surface densities are plotted as a function of the gas (i.e neglect the 24µm emission and correct for heavy elements),
surface densities to construct a K-S relation in Figure 4, using the their data points in the K-S plot are comparable to ours.
values from Table 4. We have plotted separately, the atomic and
molecular gas component of Σgas , along with the combination
of both, in red, black and blue markers respectively. Arrows de- 4. Summary and Conclusions
note the upper limits values from Table 4. In this Figure, adapted From our molecular cloud and star formation study in the tail
from Jáchym et al. (2014), we can compare our sources with with north of NGC 4388 in Virgo, we can draw the following conclu-
theirs in the Norma Cluster, as well as with the sample of spiral sions:
galaxies from Bigiel et al. (2008) (in coloured contours), and the
sample of 30 nearby galaxies from Bigiel et al. (2011), plotted 1. CO(1–0) and CO(2–1) observations were done with the
as the running medians of ΣSFR as a function of ΣH2 , and with IRAM 30-m telescope in a total of 11 targets all along the
a typical depletion time (τdep ) of ∼2.3 Gyr. Additionally, shaded ram-pressure stripped tail northeast NGC4388 in the Virgo
regions have been included to represent regions from the out- Cluster, in order to probe the presence of molecular gas un-
skirts of XUV disk galaxies. NGC4625 and NGC6946 data was der extreme conditions. Such targets were selected for having
taken from Watson et al. (2014) (priv. communication), includ- strong peaks of HI and Hα emission.
ing IRAM-30m CO observations, and Hα luminosities measured 2. Four of such positions showed CO detections, and 3 of them
within a 600 aperture. M63 (NGC5055) data corresponds to the concentrated in the HaR-2 region, at a distance of ∼70 kpc
bright UV region located at 1.35r25 in Dessauges-Zavadsky et al. of NGC4388, where molecular gas in unexpected. Given the
(2014), with IRAM 30-m CO data, and a SFR measured from the large distances of these sources to NGC4388, it is not likely
FUV and 24µm emission. that the molecular gas was stripped from the galaxy, and
“Isochrones” of constant star formation efficiencies are also must have formed in situ from the HI gas plume.
shown to indicate the depletion times of 108 , 109 and 1010 years 3. Gaussian line profiles were fitted to the spectra of the detec-
to consume all the gas, including an additional red isochrone tions, finding a wide range of velocity dispersion (∆v∼10-
to mark the age of the universe as one Hubble time (i.e τdep = 80 km/s). The CO(1–0) line profiles were used to estimate
13.8Gyr). molecular gas masses and surface densities. The amount of
Contrary to the photometry done in the HI data, the Hα data molecular gas in these 3 regions (HaR-1, HaR-2 and HaR-
was not measured in a 2200 diameter aperture, as the CO(1–0) 2-4) is very low (between 0.6 and 3 ×106 M ), and their H2
FWHM, but in an aperture of 100 in diameter, similar to the seeing surface densities between 0.8 and 2 M pc−2 . These values
of those observations. Since we are averaging this Hα emission in are well below the HI-H2 threshold, where the gas is mainly
Table 4. Kennicutt-Schmidt relation values.

R
Source ΣH2 NHI dΩa ΣHI log(LHα )b log(ΣSFR ) τdep (H2 )
[M pc−2 ] [1012 cm−2 ] [M pc−2 ] [erg s−1 ] [M yr−1 kpc−2 ] [yr]
Source-1 0.93 0.33 0.40 <35.51 < -6.19 >1.4×1012
Source-2 <0.48 0.41 0.50 <35.51 < -6.19 —
Source-3 <0.40 0.56 0.68 <35.51 < -6.19 —
Source-5 <0.48 0.60 0.69 37.51 -4.20 <7.7×109
HaR-2 1.18 2.12 2.5 37.10 -4.61 4.8×1010
HaR-3-4 <0.40 — — 37.75 -3.96 <3.6×109
HaR-2-1 <0.44 5.56 6.77 <35.51 < -6.19 —
HaR-2-2 <0.44 5.13 6.24 <35.51 < -6.19 —
HaR-2-3 <0.22 3.67 4.46 <35.51 < -6.19 —
HaR-2-4 2.09 3.08 3.74 < 35.51 < -6.19 >3.2×1012
HaR-1 0.77 2.05 2.49 35.89 -5.82 5.1×1011
Notes. Surface densities consider a solid angle Ω=2.74 kpc2 , equivalent to the CO(1–0) beamsize at the distance of the source (17.5 Mpc fom Mei
et al. 2007).
(a)
from Oosterloo & van Gorkom (2005).
(b)
from Yagi et al. (2013) and Kenney et al. (2008).
atomic and very little is know about the SFR at such low gas Chung, A., van Gorkom, J. H., Kenney, J. D. P., & Vollmer, B. 2007, ApJ, 659,
densities, hence the importance of these detections. L115
Cluver, M. E., Appleton, P. N., Boulanger, F., et al. 2010, ApJ, 710, 248
4. Using complementary data from Yagi et al. (2013) and Ken- Conselice, C. J., Gallagher, III, J. S., & Wyse, R. F. G. 2001, AJ, 122, 2281
ney et al. (2008) for Hα and from Oosterloo & van Gorkom Cortese, L., Marcillac, D., Richard, J., et al. 2007, MNRAS, 376, 157
Dasyra, K. M., Combes, F., Salomé, P., & Braine, J. 2012, A&A, 540, A112
(2005) for HI, we computed ΣSFR and ΣHI to plot , in com- Davies, J. I., Bianchi, S., Cortese, L., et al. 2012, MNRAS, 419, 3505
DeMaio, T., Gonzalez, A. H., Zabludoff, A., Zaritsky, D., & Bradač, M. 2015,
bination with ΣH2 , a K-S relation. Our sources show an ex- MNRAS, 448, 1162
tremely low SFR (up to 2 order of magnitude lower than for Dessauges-Zavadsky, M., Verdugo, C., Combes, F., & Pfenniger, D. 2014, A&A,
typical spiral galaxies). For example, HaR-1 and HaR-2 have 566, A147
Edge, A. C., Oonk, J. B. R., Mittal, R., et al. 2010, A&A, 518, L46
total gas depletion times of 2.1×1012 and 1.5×1011 years re- Fabian, A. C., Sanders, J. S., Taylor, G. B., et al. 2006, MNRAS, 366, 417
spectively. If we consider just the molecular gas component, Feldmeier, J. J., Mihos, J. C., Morrison, H. L., Rodney, S. A., & Harding, P.
2002, ApJ, 575, 779
these depletion times are 5.1×1011 and 4.8×1010 years. Fur- Ferland, G. J., Fabian, A. C., Hatch, N. A., et al. 2009, MNRAS, 392, 1475
thermore, Source-1 and HaR-2-4 have H2 depletion times Gavazzi, G., Boselli, A., Mayer, L., et al. 2001, ApJ, 563, L23
Gu, L., Yagi, M., Nakazawa, K., et al. 2013, ApJ, 777, L36
even greater than 1.4 and 3.1 ×1012 years respectively. These Gunn, J. E. & Gott, III, J. R. 1972, ApJ, 176, 1
high values of depletion times exceed by far a Hubble time, Iwasawa, K., Wilson, A. S., Fabian, A. C., & Young, A. J. 2003, MNRAS, 345,
thus indicating that this molecular gas will not form stars 369
Jáchym, P., Combes, F., Cortese, L., Sun, M., & Kenney, J. D. P. 2014, ApJ, 792,
eventually, but remain in a gaseous phase and later join the 11
ICM. Jáchym, P., Palouš, J., Köppen, J., & Combes, F. 2007, A&A, 472, 5
Kenney, J. D. P., Tal, T., Crowl, H. H., Feldmeier, J., & Jacoby, G. H. 2008, ApJ,
5. From Figure 4 we can see the linearity between the SFR and 687, L69
the gas surface density, so clear at high gas surface densities Kenney, J. D. P., van Gorkom, J. H., & Vollmer, B. 2004, AJ, 127, 3361
Kenney, J. D. P. & Young, J. S. 1989, ApJ, 344, 171
(> 9 M pc−2 ) for normal spiral galaxies, cannot be extrap- Kennicutt, R. C. & Evans, N. J. 2012, ARA&A, 50, 531
olated to lower densities, below the HI-H2 threshold, where Leroy, A. K., Walter, F., Sandstrom, K., et al. 2013, AJ, 146, 19
Lim, J., Ohyama, Y., Chi-Hung, Y., Dinh-V-Trung, & Shiang-Yu, W. 2012, ApJ,
the star formation is extremely inefficient, and the molecular 744, 112
gas, even though present, does not necessarily form stars . Machacek, M., Dosaj, A., Forman, W., et al. 2005, ApJ, 621, 663
Machacek, M. E., Jones, C., & Forman, W. R. 2004, ApJ, 610, 183
Mahdavi, A., Geller, M. J., Fabricant, D. G., et al. 1996, AJ, 111, 64
Acknowledgements. The IRAM staff is gratefully acknowledged for their help Mei, S., Blakeslee, J. P., Côté, P., et al. 2007, ApJ, 655, 144
in the data acquisition. We thank J. Kenney for facilitating important Hα data. Merritt, D. 1984, ApJ, 276, 26
Mihos, J. C., Harding, P., Feldmeier, J., & Morrison, H. 2005, ApJ, 631, L41
F.C. acknowledges the European Research Council for the Advanced Grant Pro- Nulsen, P. E. J. 1982, MNRAS, 198, 1007
gram Number 267399-Momentum. We made use of the NASA/IPAC Extragalac- Oosterloo, T. & van Gorkom, J. 2005, A&A, 437, L19
tic Database (NED), and of the HyperLeda database. C.V. acknowledges finan- O’Sullivan, E., Giacintucci, S., Vrtilek, J. M., Raychaudhury, S., & David, L. P.
cial support from CNRS and CONICYT through agreement signed on December 2009, ApJ, 701, 1560
11th 2007. Quilis, V., Moore, B., & Bower, R. 2000, Science, 288, 1617
Roediger, E. & Hensler, G. 2005, A&A, 433, 875
Salomé, P., Combes, F., Edge, A. C., et al. 2006, A&A, 454, 437
Salomé, P., Combes, F., Revaz, Y., et al. 2011, A&A, 531, A85
Scott, T. C., Cortese, L., Brinks, E., et al. 2012, MNRAS, 419, L19
Serra, P., Koribalski, B., Duc, P.-A., et al. 2013, MNRAS, 428, 370
References Sun, M., Donahue, M., Roediger, E., et al. 2010, ApJ, 708, 946
Sun, M., Donahue, M., & Voit, G. M. 2007, ApJ, 671, 190
Bigiel, F., Leroy, A., Walter, F., et al. 2008, AJ, 136, 2846 Tamura, T., Maeda, Y., Mitsuda, K., et al. 2009, ApJ, 705, L62
Tonnesen, S., Bryan, G. L., & van Gorkom, J. H. 2007, ApJ, 671, 1434
Bigiel, F., Leroy, A. K., Walter, F., et al. 2011, ApJ, 730, L13 Vollmer, B., Beckert, T., & Davies, R. I. 2008, A&A, 491, 441
Boissier, S., Boselli, A., Duc, P.-A., et al. 2012, A&A, 545, A142 Vollmer, B., Braine, J., Combes, F., & Sofue, Y. 2005, A&A, 441, 473
Boissier, S., Gil de Paz, A., Boselli, A., et al. 2008, ApJ, 681, 244 Vollmer, B., Cayatte, V., Balkowski, C., & Duschl, W. J. 2001, ApJ, 561, 708
Watson, L. C., Martini, P., Lisenfeld, U., et al. 2014, in American Astronomical
Braine, J., Lisenfeld, U., Due, P.-A., & Leon, S. 2000, Nature, 403, 867 Society Meeting Abstracts, Vol. 223, American Astronomical Society Meet-
Casasola, V., Hunt, L., Combes, F., & Garcia-Burillo, S. 2015, ArXiv e-prints ing Abstracts 223, 454.22
[arXiv:1503.00280] Yagi, M., Gu, L., Fujita, Y., et al. 2013, ApJ, 778, 91
Cayatte, V., van Gorkom, J. H., Balkowski, C., & Kotanyi, C. 1990, AJ, 100, 604 Yagi, M., Komiyama, Y., Yoshida, M., et al. 2007, ApJ, 660, 1209
Yoshida, M., Yagi, M., Okamura, S., et al. 2002, ApJ, 567, 118
Chung, A., van Gorkom, J. H., Kenney, J. D. P., Crowl, H., & Vollmer, B. 2009, Zhang, B., Sun, M., Ji, L., et al. 2013, ApJ, 777, 122
AJ, 138, 1741

Fig. 4. K-S relation for sources in Table 4, with filled circles and arrows for values and upper limits respectively. Red markers consider only H2
for Σgas , black ones only HI, and blue ones the sum of both. Figure adapted from Jáchym et al. (2014), where their sources ESO137-001, 001-A,
001-B and 001-C have been plotted in a similar way than ours: red circles for H2 gas, black left arrows for HI upper limits, and blue "error bars"
to account for both gas components. Colored contours account for the spiral galaxies from Bigiel et al. (2008) (green, orange, red and purple for
1, 2, 5 and 10 sampling points per 0.05 dex respectively). Black markers with error bars correspond to the running medians in ΣS FR as a function
of σH2 of 30 nearby galaxies from Bigiel et al. (2011). Shaded ovals represent the data from the outer parts of XUV disk galaxies: NGC4625 and
NGC6946 from Watson et al. (2014) (priv. communication) , and M63 (NGC5055) from Dessauges-Zavadsky et al. (2014), taking only H2 into
account in all of them. The dashed vertical line shows the 9 M pc−2 threshold at which the atomic gas saturates. Dashed inclined lines represent
“isochrones” of constant star formation efficiencies, indicating the depletion times τdep of 108 , 109 and 1010 years to consume all the gas. The
dashed red isochrone marks a depletion time equal to the age of the universe, as one Hubble time (13.8 Gyr). A representative shift of the HaR-1
marker for ΣH2 is drawn, to show the "effective" molecular gas density at which stars would formed in this region if we consider a beam correction
factor of +2.68 in log space, to convert our 2200 beam to a ∼100 beam as in the Hα data.

Chapter 4
Summary and Conclusions
This thesis presented the results of three years of work dedicated to obtain the
degree of Doctor in Astronomy and Astrophysics of the Observatoire de Paris. It
focused on star formation in low gas density and low metallicity environments, par-
ticularly in the search of carbon monoxide (CO) as a tracer of molecular gas (H2 ),
which is the main precursor of star formation. Due to the extreme conditions of such
environments, the detection of CO in these places is very difficult, but plausible. In
fact its presence is denoted by different signs of (scarce) star formation activity.
The importance of the regions studied in this thesis not only relies on determining
constrains on the environmental conditions needed to form stars, but also on the
resemblance of such regions with a younger universe, much hotter, with lower levels
of gas density and content of heavy elements.
This was mainly an observational work, with CO observations at the IRAM 30m
radio-telescope, and with complementary data of space and ground-based telescopes,
to trace different stages of the star formation process. All this data was used to
quantify the star formation activity and its efficiency, through Kennicutt-Schmidt
(K-S) relations, which compares the SFR surface density versus the gas surface
density. The goals of this work were to find molecular gas in places were it is
unexpected due to hostile conditions, and use it, along with the complementary
data, to characterize the star formation efficiency of the process, which is very
unexplored at such low gas density regions.
XUV galaxies
The first topic explored in this thesis was on nearby XUV galaxies, i.e, galaxies
with extended disks of UV emission, first revealed by the GALEX telescope (Gil de
Paz et al. 2007a). Such disks extend up to the most external parts of spiral galaxies,
beyond their optical radius (r25 ), where gas densities and metallicities are low (Henry
& Worthey 1999). Many of these regions lack Hα emission (e.g Thilker et al. 2005),
the most common star formation tracer, but are rich in NUV and FUV emission
coming from young O and B stars, indicating recent episodes of star formation, and
102
the likely presence of the molecular gas out of which they must have been formed.
This thesis began in 2012 by resuming a work started in 2007 with CO obser-
vations on the nearby M63 galaxy, which has a remarkable disk of UV emission,
extending up to 2.5 times its optical radius, as well as an extended disk of HI emis-
sion (Battaglia et al. 2006). Simultaneous CO(1–0) and CO(2–1) observations were
done with the IRAM 30m telescope, using a linear arrange of 27 single ON+OFF
pointings placed along the major axis of the galaxy, from the center of the galaxy
up to a galactic radius (rgal ) of 1.6r25 , and 12 pointings arranged in a rectangular
configuration to map a small region bright in UV, located at rgal = 1.36r25 . This
mapping technique probed to be very efficient for detecting faint CO and achieving
a good S/N. CO(1–0) emission was found in the observations along the radial axis,
from the center of the galaxy up to the optical radius r25 , but not beyond. However,
a CO(1–0) detection was found in the external bright UV region at 1.36r25 . This is
the fourth molecular gas detection in the outskirts of nearby spiral galaxies (after
NGC4414 by Braine & Herpin 2004, NGC6946 by Braine et al. 2007 and M33 by
Braine et al. 2010). The CO(2–1) emission is, on the other hand, confined to a
galactic radius of 0.68r25 . This result suggests subthermal excitation in the outer
regions of the M63 disk.
A set of complementary data was used, including HI, Hα, 24µm, NUV, FUV and
public CO, to compare the radial profiles of all this tracers to our CO observations.
These radial profiles show a severe drop of emission with galactic radius, basically for
all tracers except HI. Close to the r25 limit, all tracers begin to vanish considerably,
and beyond r25 , they all are practically absent with the exception of the faint UV and
HI emission. This is particularly seen at the location of the bright UV region, where
the CO(1–0) detection was found, and where UV and HI emission, even though
faint, still exist. This shows how UV and HI emission are good tracers to look for
molecular gas (e.g Thilker et al. 2005, Gil de Paz et al. 2005, 2007a, Crosthwaite
et al. 2002, Nieten et al. 2006). The UV region at rgal = 1.36r25 , in which the CO
emission is detected, is characterized by FUV and NUV emission fluxes similar to the
fluxes observed at r25 , but stronger than the fluxes observed at rgal > r25 along the
M63 major axis. This probably reflects a tight correlation between the CO and UV
fluxes, namely between the intensity of star formation and the amount of molecular
gas, so it strongly suggests that the absence of CO detection at rgal > r25 along
the radial axis, where the XUV is weaker, is simply the result of the CO detection
threshold that is still too high.
Additionally, the external UV region is characterized by a very high HI flux with
respect to the measured CO flux. This leads us to speculate that HI is more likely
the precursor of H2 rather than the product of UV photodissociation, since it seems
to dominate in quantity. This is, however, true as long as substantial H2 is not
hidden in the outer disk regions of M63, which may be the case as we observe hints
for an excitation temperature decrease at large galactocentric radii, that could lead
to very weak CO lines
SFRs were calculated for each pointing position of our observations, using the
FUV and 24µm complementary data. In the same way, gas surface densities were
103
calculated, using our observations to obtain ΣH2 , and the HI complementary data to
calculate ΣHI , as well as the combination of both (ΣHI+H2 ). These results were used
to construct a K-S relation including all observing pointings in M63, i.e the ones
along the major axis and the ones covering the external bright UV region. This is
the first time that the K-S relation is quantified in the outskirts of a spiral galaxy,
i.e., in low gas density environments, and we see that at a molecular gas surface
density as low as ΣH2 = 0.35M pc2 . This value is well below all the determined H2
surface densities referenced in spiral galaxies so far (Bigiel et al. 2008, 2011), and
still star formation occurs spontaneously.
Schmidt (1959) was the first to characterize the SFR density as a function of
the gas density (both initially in volumic densities but soon translated into surface
densities for observational reasons) through a power law of the form ΣSFR = AΣN gas ,
with the coefficient A directly related to the SFE, and the exponent N an indicator of
the linearity of the relation. For the K-S relation obtained here, linear regressions of
the form log(ΣSFR ) = log A + N log(Σgas ) were fitted to both data sets individually
(radial profile and bright UV region), obtaining values of A and N for each one of
them. We see a clear separation of these two sets of observations in the K-S plot,
with the data points of major axis located above ∼10 M pc−2 (the HI-H2 threshold),
and the points of the bright UV region located below this limit (Figure 2.12). We see
that pointings along the major axis have their gas surface densities highly dominated
by molecular gas, and follow an almost linear relation (N∼1, in log space) when
molecular gas is considered for Σgas . On the other hand, pointings corresponding
to the bright UV region have their gas surface densities mainly dominated by the
atomic gas component, and follow a very non-linear (N∼4) K-S relation. A clear
broken power law is visible in the K-S plot between the two data sets, when the total
gas surface density is considered (ΣHI+H2 ), with an N slope of 1.7 for the pointings
along the major axis, and 4.7 for the ones in the bright UV region. This discontinuity
in the K-S relation indicates that the star forming processes, so well studied in the
inner regions of galaxies, where gas densities are higher, are not the same, in terms
of efficiency, to the ones in the less explored outer regions, with lower gas densities,
and the relation cannot be extrapolated from inner to outer parts of galaxies.
This discontinuity can also be appreciated directly in the differences in SFE.
Indeed, along the major axis of the disk out to the isophotal radius, the SFE settles
the gas consumption within 108 years from 1% to 10%, whereas in the external
UV region much less than 1% of the gas is converted into stars within 108 years.
Consequently, star formation still occurs in the outer regions of the disk, but at very
low efficiency, perhaps owing to the flaring of the outer gas layers.
Motivated by the successful results obtained for M63, new CO observations were
done in a set of nearby galaxies with extended UV disks. Three XUV galaxies were
selected for having extended disks of both UV and HI emission, well beyond their
optical limits. ON+OFF single observing pointings were done in several targets
with peaks of UV and HI emission, beyond the optical limits. Both lines, CO(1–0)
and CO(2–1) were observed simultaneously with the IRAM 30m telescope, but no
104
clear detection was found, probably because of a considerable loss of observing time
due to bad weather conditions (strong winds). Weak emission was even searched for
through a stacking technique, which enables us to increase the S/N of the data by
averaging the spectra of all targets observed, at the same velocity (v=0). This is done
via a velocity shift in each one of the spectra using HI complementary information.
Unfortunately, no detection was found either, but faint emission is not completely
discarded yet, and new search techniques and more elaborated stacking procedures
will be applied in the near future to search for very faint emission. Nevertheless, 3σ
upper limits were calculated from the rms noise levels of the spectra.
Finally, the last XUV galaxy studied was M83, which is a nearby and face-on
galaxy, with an extended disk of both UV (Gil de Paz et al. 2007b) and HI (Walter
et al. 2008) emisison. New ALMA CO(2–1) and continuum observations were done
in an external region of the M83 galaxy, beyond its optical radius, in a small region
very bright in both UV and HI (Figure 2.18). A 300 ×1.500 region was mapped within
just an hour of observation, reaching an rms of 10.3 mJy per channel and a spectral
resolution of 2.5 km/s. Unfortunately, apart from a very weak point source in
continuum, no clear CO detection has been found yet. It is possible that no CO
emission is present due to an extremely low metallicity and gas density, or possibly
due to gas flaring in this region, so far away from the center of the galaxy, and
so vulnerable to external factors. However, due to the recent delivery of the data,
this is still work in progress. There seems to be some hints of faint clumps (Figure
2.20), with possible weak emission, and will be analyzed in a near future, with
more elaborated search algorithms. Nevertheless, star formation activity is present.
Maps of SFR surface density were created for the observed region, with FUV and
24µm complementary data, finding a ΣSFR distribution mainly dominated by the
24µm emission (which traces the obscured FUV emission that is re-emitted by dust
grains), and with values as high as 2.9×10−3 M yr−1 kpc−2 .
Ram-pressure stripping
The second part of this thesis was dedicated to investigate the effects of ram-
pressure striping as a possible trigger of star formation, in regions far away from
galaxies, with extremely low levels of gas density and metallicity, and where no star
formation activity is expected.
The Virgo Cluster is an ideal laboratory to study this effect, since it harbors
more than a thousand spiral and elliptical galaxies that interact with each other
through merging and collisional processes, and with the cluster as a whole and the
ICM through tidal interactions. For this particular work, the NGC4388 galaxy was
chosen in the Virgo Cluster, for having a long tail of atomic gas (Oosterloo & van
Gorkom 2005) stripped from the galaxy by the ICM. The HI tail extends more than
100 kpc from NGC4388 in direction to the M86 galaxy, and has previous detections
of Hα (Yagi et al. 2013) and X-ray emission (Iwasawa et al. 2003).
Several observing targets where selected all along the tail, for having peaks of HI
and Hα emission (from complementary data). Simultaneous CO(1–0) and CO(2–
105
1) observations were done with the IRAM 30m telescope, successfully finding 3
detections in a region which is bright in both HI and Hα emission, and located ∼70
kpc away from NGC4388. Due to the high distance to the galaxy, and because
ram-pressure is not expected to be able to drag such a dense gas, especially so far
away, an H2 formation in situ is most reasonable for this molecular gas.
For the sources with CO detections, a wide range of velocity dispersion (∆v ∼
10 − 80 km/s) was found. From Gaussian line fitting, CO(1–0) line profiles were
obtained and used to estimate molecular gas masses and surface densities. The
amount of molecular gas in these regions is very low (between 0.6 and 3 ×106 M ),
and their H2 surface densities between 0.8 and 2 M pc2 . These values are well below
the HI-H2 threshold, where the gas is mainly atomic and very little is known about
the SFR at such low gas densities, hence the importance of these detections. For
the remaining sources with no visible detection, 3σ upper limits where calculated
from the rms noise levels of the spectra.
For the observed regions, ΣSFR and ΣHI where calculated from complementary Hα
and HI data respectively, as well as ΣH2 from our CO observations, to construct a K-
S relation, including detections and upper limits. All these regions lay on the low gas
density portion of the K-S plot, below the HI-H2 threshold, and show an extremely
low SFR (up to 2 orders of magnitude lower than for typical spiral galaxies). In the
K-S plot, they are only comparable with the most outer clumps in the Hα/X-ray tail
of the ISM stripped galaxy ESO137-001 in the Norma Cluster (Jáchym et al. 2014)
and the XUV disk galaxies from Watson et al. (2014) and Dessauges-Zavadsky et al.
(2014).
The inefficiency of the star forming processes in such regions can be directly seen
in the extremely high depletion times, with some values even greater than 1012 years,
exceeding by far a Hubble time. This suggests that this molecular gas will never be
able to form stars, but remain in a gaseous phase and later join the ICM. Finally,
we see that the observed sources follow a non-linear regime in the K-S plot. The
linearity between the SFR and the gas surface density, so clear at high gas surface
densities (> 9 M pc−2 ) for normal spiral galaxies, once again cannot be extrapolated
to lower densities, below the HI-H2 threshold, where the star formation is extremely
inefficient, and the molecular gas, even though present, does not necessarily form
stars.
We conclude that CO observations in extremely low gas density regions can suc-
cessfully trace molecular gas, when atomic gas and star formation tracers, such as
UV, IR or Hα, are present. Furthermore, we see that H2 formation in situ is possi-
ble, even in the most hostile environments. The detection of molecular gas helps us
not only to confirm ongoing star formation, but allows us to quantify it and com-
pare it to the star formation processes taking place in other regions, with different
environmental conditions. Through analyses of Kennicutt-Schmidt relations, we see
106
that star formation is different depending on the regions and conditions involved,
since they show different SFEs. The aim of this thesis is to make a contribution to
the less explored field of star formation under low density conditions. Thanks to an
extended observational work, this research was able to give not only a quantification
of star formation in the outer parts of spiral galaxies for the first time, but also helps
to support the idea that star formation processes in low gas density environments are
not the same as the ones at higher densities. Therefore, the Kennicutt-Schmidt rela-
tion and the star forming processes, so well studied at densities above ∼10 M pc−2 ,
cannot be extrapolated to lower densities.
107
Appendix
A.1 CO in low metallicity galaxies

This appendix includes an additional work on star formation at low metallicity
environments, particularly in the WLM galaxy, with the first CO ever on a galaxy
with such a low metallicity. I started to work in this collaboration during my master
thesis, and finished during my first year of PhD.
A.1.1 Introduction
Searching for molecular gas in low metallicity galaxies always has an observational
restriction due to its low emission. At low metallicity, any other element that is not
hydrogen or helium is sparse, including carbon and oxygen, or any other elements
that can trace molecular gas. In fact, before this work, CO had never been detected
in galaxies in which the oxygen abundance relative to hydrogen is less than 20% of
that of the Sun. But these galaxies do not lack completely of star formation, since
there’s evidence that they host young stars.
Wolf-Lundmark-Melotte (WLM) (Figure 1) is an isolated dwarf galaxy at a dis-
tance of 985 kpc, at the edge of the Local Group. It is a small, dwarf irregular
(dIrr) galaxy with a metallicity of 0.13Z , a low SFR due to its size, and no pre-
vious evidence of the molecular gas that always accompanies young stars in larger
galaxies.
Star forming gas was searched through CO(3–2) emission and continuum dust
emission at 345 GHz with the Atacama Pathfinder Experiment (APEX) telescope
in Chile. Both the Swedish Heterodyne Facility Instrument (SHeFI) and the APEX
Bolometer Camera (LABOCA) were used, and the targets were selected for having
peaks of 160µm (SPITZER) and HI (VLA) emission. The reason for choosing the
band at 345 GHz is because dust measurements can be converted to a dust tem-
perature and a dust mass, and, after applying a suitable gas-to-dust ratio, to a gas
mass from which the H I mass can be subtracted to give the H2 mass for comparison
with CO.
108
Figure 1: RGB image of WLM. Red: HI, green: V band, blue: FUV. Black circles denote the two
positions where CO(3–2) was found: A position to the upper right, and B position to the lower
left.
109
A.1.2 Results
Figure 1 shows WLM and the two regions, designated A and B, where CO(3–2)
emission was detected. The peak CO brightness temperature in each detected region
is ∼0.01 - 0.015 K and the linewidth is 12 km s−1 (FWHM). Our observations with
an 1800 aperture yields an intensity of 0.200 ± 0.046 K km s−1 and a luminosity of
1500 K km s−1 pc2 for CO(3–2) in region A.
A large HI and FIR cloud that surrounds region A, designated region A1, was
used to measure the dust temperature, Td = 15 K, which was assumed to be the same
throughout the region (the 160µm observation does not resolve region A, and so a
more localized temperature measurement is not possible). This dust temperature
was determined from the continuum emission, using the 870µm and 160µm fluxes,
corrected for the CO(3-2) line and broadband free–free emission, to construct spec-
tral energy distributions (SED). A modified black-body function was fitted to the
SED, with dust emmissivity proportional to frequency to the power of the spectral
emmissivity index β:
Sν ∝ ν β Bν (Td ) (1)
Local measurements (Planck Collaboration et al. 2011) suggest that β = 1.78 ±

0.08, although a range is possible (Draine 2003, Galametz et al. 2012), depending
on grain temperature and properties (Coupeaud et al. 2011). The 870µm flux was
also corrected for an unexplained FIR and sub-millimeter excess that is commonly
observed in other low-metallicity galaxies (Galametz et al. 2011, Verdugo 2012)
We made a χ2 minimization fitting with the relation in Equation 1, using β=1.8
leaving Td as the free paramenter. Additionally, we repeated the fitting with other
values of β to show the dependency of the results on the spectral emmissivity index
(Figure 2 shows an test example of such procedure).
Using this dust temperature for region A, the dust mass was calculated from an
emissivity (Draine 2003) of κ = 13.9 cm2 gm−1 at 140µm, and converted to 870µm
with the same power-law index, β. The dust mass for region A is then
S870,A D2
Mdust,A = (2)
κBν (Td )
for flux S870,A , distance D and black-body spectral function Bν .

Dust mass is converted to gas mass using a factor equal to the gas–to–dust mass
ratio, RGD . An approximation (Leroy et al. 2011) is to assume the solar value
(Draine et al. 2007) (1/0.007) increased by the inverse of the metallicity of WLM
(0.13), which would give 1100. We use this approximation here, but introduce
a scaling factor to the gas mass, δGD = RGD /1100; that is, the gas-to-dust ratio
normalized by the solar value and scaled to the metallicity.
110
Figure 2: Example of a modified blackbody fitting procedure for region A in a 6000 aperture.
Different values of β where tested. Each fit gives a result for Td .
The total gas mass column density in a 2200 region around region A is ∼58 ± 15
M pc−2 for β = 1.8. Atomic hydrogen contributes ∼27.3 ± 1.4 M pc−2 , and the
remainder is ascribed to molecular H2 traced by the observed CO.
The integral under the CO(3–2) line from region A is ICO = 0.200 ± 0.046 K
km s−1 . This intensity has to be converted to CO(1–0) before comparing it with H2
mass in the conventional way. We take as a guide (Nikolić et al. 2007) the value of
CO(3–2)/CO(1–0) ∼ 0.80 in another low metallicity galaxy, the SMC (Dufour 1984)
(where O/H = 20% of the solar value). The result, 0.25 K km s−1 , is combined with
the H2 mass column density to determine the conversion factor, αCO , from CO(1–0)
to H2 .
Taking the H2 column density from the residual between the dust-derived total
and the HI column density, 31 ± 15 M pc−2 , and dividing by the inferred CO(1–
0) line integral of 0.25 K km s−1 , we obtain αCO = 124 ± 60 M pc−2 K−1 km−1 s
including He and heavy elements, with a range in αCO from 34 to 271 as β varies
from 1.6 to 2. The corresponding CO–to–H2 conversion factor is XCO = (5.8 ±
2.8)×1021 cm−2 K−1 km−1 s. This value is used to calculate the H2 mass in region B
from its CO line intensity, giving a value of (1.2 ± 0.6) ×105 M .
The star-formation rate based on the Hα and FUV (Hunter et al. 2010) fluxes
within an 1800 aperture centered on cloud A is (3.9–4.8) × 10−5 M yr−1 . Divid-
ing these rates into the CO-associated molecular mass using αCO = 124 M pc−2
K−1 km−1 s gives a CO molecular consumption time (for converting gas into stars)
of 4.6–3.8 Gyr for region A. In region B, the star-formation rate from Hα and FUV
fluxes is (1.7–12.6) × 10−5 M yr−1 and the CO molecular consumption time is 6.7–
1.5 Gyr. These times are only slightly larger than the average value in spiral galaxies
(Leroy et al. 2008), which is 2 Gyr, but they are ten times larger than the rate per
molecule in local giant molecular clouds (Lada et al. 2012), which is a more direct
111
analogy with our observations. The detection of CO in WLM suggests that star
formation continues to occur in dense molecular gas even at lower metallicities than
previously observed. The similarity between the metallicities of dIrr galaxies such
as WLM and those of larger galaxies at high redshift (Mannucci et al. 2009) implies
that we should be able to study star formation in young galaxies using the usual
techniques.
A.1.3 Publication
This entirety of this work was published by Nature on March 2013, and is pre-
sented here.
112
LETTER doi:10.1038/nature11933
Carbon monoxide in clouds at low metallicity in the

dwarf irregular galaxy WLM
Bruce G. Elmegreen1, Monica Rubio2, Deidre A. Hunter3, Celia Verdugo2, Elias Brinks4 & Andreas Schruba5
Carbon monoxide (CO) is the primary tracer for interstellar clouds dust measurements can be converted to a dust temperature and a dust
where stars form, but it has never been detected in galaxies in which mass, and, after applying a suitable gas-to-dust ratio, to a gas mass
the oxygen abundance relative to hydrogen is less than 20 per cent from which the H I mass can be subtracted to give the H2 mass for
of that of the Sun, even though such ‘low-metallicity’ galaxies often comparison with CO.
form stars. This raises the question of whether stars can form in Figure 1 shows WLM and the two regions, designated A and B,
dense gas without molecules, cooling to the required near-zero where we detected CO(3–2) emission, along with H I, far-infrared
temperatures by atomic transitions and dust radiation rather than (FIR) and submillimetre images. Observed and derived parameters
by molecular line emission1; and it highlights uncertainties about are listed in Tables 1 and 2, respectively. The peak CO brightness
star formation in the early Universe, when the metallicity was temperature in each detected region is ,0.01–0.015 K and the line-
generally low. Here we report the detection of CO in two regions width is ,12 km s21 (full-width at half-maximum). Previous efforts to
of a local dwarf irregular galaxy, WLM, where the metallicity is 13 detect CO(J 5 1–0) in WLM5 partly overlapped region A with a 4599
per cent of the solar value2,3. We use new submillimetre observa- aperture and determined a 5s upper limit to the CO(1–0) intensity of
tions and archival far-infrared observations to estimate the cloud 0.18 K km s21. Our observation with an 1899 aperture yields an intens-
masses, which are both slightly greater than 100,000 solar masses. ity of 0.200 6 0.046 K km s21 for CO(3–2) in the same region. The
The clouds have produced stars at a rate per molecule equal to 10 difference arises because the CO cloud is unresolved even by our
per cent of that in the local Orion nebula cloud. The CO fraction of 1899 beam—we did not detect comparable CO(3–2) intensities in
the molecular gas is also low, about 3 per cent of the Milky Way our searches adjacent to region A. The previous upper limit corre-
value. These results suggest that in small galaxies both star-form- sponds to a maximum CO(1–0) luminosity of 8,300 K km s21 pc2
ing cores and CO molecules become increasingly rare in molecular inside 4599 (which corresponds to a beam diameter of 215 pc at
hydrogen clouds as the metallicity decreases. WLM), whereas the cloud we detect has a CO(3–2) luminosity ,6
Wolf–Lundmark–Melotte (WLM) is an isolated dwarf galaxy at the times smaller (1,500 K km s21 pc2). Likewise, the previous null detec-
edge of the Local Group4. It has a low star-formation rate because of its tion5 in CO(J 5 2–1) claimed a 5s upper limit that is about the same as
small size and, like other dwarf irregular (dIrr) galaxies, shows no our CO(3–2) detection, but their closest pointing differed from
previous evidence5 for the molecular gas that always accompanies region A by ,70 pc (1499, or half the beam diameter for CO(2–1)),
young stars in larger galaxies6. One problem with the detection of which could have been enough to take it off the CO cloud.
molecules is that the dominant tracer of such gas is CO, and dIrr The 160-mm, 870-mm and H I peaks are slightly offset from the CO
galaxies have low carbon and oxygen abundances relative to hydrogen. positions, indicating variations in temperature and molecular frac-
No galaxy with an O/H abundance less than 20% has been detected tion. A large H I and FIR cloud that surrounds region A, designated
using CO as a tracer7–9. Far more abundant is molecular hydrogen region A1, was used to measure the dust temperature, Td , 15 K, which
(H2), but this does not have an observable state of excitation at the was assumed to be the same throughout the region (the 160-mm obser-
low temperatures (,10–30 K) required for star formation. vation does not resolve region A, and so a more localized temperature
To search for star-forming gas, we surveyed WLM for CO(J 5 3–2) measurement is not possible). We determined Td from the 870-mm and
emission in rotational state J and for continuum dust emission at 160-mm fluxes corrected for the CO(3–2) line and broadband free–free
345 GHz using the Atacama Pathfinder Experiment (APEX) telescope emission (Table 1), assuming a modified black-body function with dust
at Llano de Chajnantor, Chile, with the Swedish Heterodyne Facility emissivity proportional to frequency to the power b. Local measure-
Instrument10 and the Large APEX Bolometer Camera11 (LABOCA). ments13 suggest that b 5 1.78 6 0.08, although a range is possible14,15,
We also used a map of dust emission at 160 mm from the Spitzer Local depending on grain temperature and properties16. The 870-mm flux was
Volume Legacy Survey12 and a map of atomic hydrogen re-reduced also corrected for an unexplained FIR and submillimetre excess that is
from the archives of the Jansky Very Large Array radio telescope. The commonly observed in other low-metallicity galaxies17,18. An alternate
Table 1 | Observations of WLM
Source Region Right ascension Declination Beam diameter (99) Flux
CO(3–2) A 0 h 1 min 57.32 s 215u 269 49.599 18 0.200 6 0.046 K km s21

HI A 0 h 1 min 57.32 s 215u 269 49.599 22 774 6 40 mJy km s21
870 mm A 0 h 1 min 57.32 s 215u 269 49.599 22 2.66 6 0.53 mJy (0.11, 0.02)*
HI A1 0 h 1 min 56.93 s 215u 269 40.8499 45 4,170 6 82 mJy km s21
870 mm A1 0 h 1 min 56.93 s 215u 269 40.8499 45 15.2 6 3.0 mJy (0.11, 0.06)*
160 mm A1 0 h 1 min 56.93 s 215u 269 40.8499 45 136.2 6 13.6 mJy (0.05){
CO(3–2) B 0 h 2 min 1.68 s 215u 279 52.599 18 0.129 6 0.032 K km s21
* Quantities in parentheses are the CO(3–2) flux and the free–free emission, both in mJy, that were subtracted from the source flux before calculating the dust flux.
{ Quantity in parentheses is the free–free emission, in mJy, that was subtracted from the source flux before calculating the dust flux. The average FIR excess factor18 for the Small Magellanic Cloud (SMC) is 1.7, so we
divide the CO-corrected and free–free-corrected 870-mm fluxes in the table by 1.7 to get the thermal dust flux.
1
IBM Research Division, T.J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights, New York 10598, USA. 2Departamento de Astronomı́a, Universidad de Chile, Casilla 36-D, Santiago, Chile.
3
Lowell Observatory, 1400 West Mars Hill Road, Flagstaff, Arizona 86001, USA. 4Centre for Astrophysics Research, University of Hertfordshire, Hatfield AL10 9AB, UK. 5Cahill Center for Astronomy and
Astrophysics, California Institute of Technology, Pasadena, California 91125, USA.
2 8 M A R C H 2 0 1 3 | VO L 4 9 5 | N AT U R E | 4 8 7
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH LETTER
a b 0.06 Figure 1 | Observations of the

Region A CO galaxy WLM. WLM is a small, gas-
HI rich galaxy 985 6 33 kpc from the
CO, Tmb (K); H I, S (Jy)

0.04 Milky Way4. It contains
1.6 3 107M8 of stars28, compared
with (6.4 6 0.6) 3 1010M8 in the
0.02
Milky Way29, and it forms new stars
at a rate23 of 0.006M8 yr21, which is
0 12 times higher per unit stellar mass
A than the Milky Way30. a, Colour
–200 –150 –100 –50 composite image: red, H I; green,
V band; blue, GALEX far-ultraviolet.
B 0.06 For H I the aperture was 7.699 and the
Region B
resolution was 2.6 km s21, and for
CO, Tmb (K); H I, S (Jy)

0.04
CO(3–2) the aperture was 1899
(circles) and the resolution was
0.11 km s21, although the CO(3–2)
0.02 spectra shown in the figure were
smoothed to a resolution of
0 2.2 km s21. The CO detections are
labelled; their exposure times were
218 min (region A) and 248 min
–200 –150 –100 –50 (region B). Other regions searched
Velocity (km s–1) with exposure times shorter by
factors of ,2 to ,6 are indicated by
c d plus signs; the presence of
comparable CO mass in some of
these other regions cannot be ruled
out. b, Spectra of the two detections:
velocities are relative to the local
A standard of rest; CO labels main-
A
beam brightness temperature, Tmb,
in kelvin; H I labels flux in Jy. c, False-
B colour image of the 870-mm
B
observations made with LABOCA on
APEX. d, False-colour Spitzer 160-
mm image obtained from Spitzer
archives. In c and d, the images show
the same field of view and the small
circles (2299 diameter, the resolution
of LABOCA) show where CO was
60′′ detected. The large circle is 4599 in
60′′ diameter and surrounds a large H I
and FIR cloud (region A1) where the
dust temperature was measured.
Table 2 | Derived quantities for WLM

Source Region T (K) S* (M8 pc22) Mass (M8 )
b 5 1.8, aCO 5 124 6 60{

Dust A 14.7 6 0.7{ 0.053 6 0.014 (4.6 6 1.2) 3 102
Gas1 A (58 6 15)dGD ((5.1 6 1.3) 3 105)dGD
H || A 27.3 6 1.4 (2.4 6 0.1) 3 105
H2 A 31 6 15 (1.8 6 0.8) 3 105
H2 " B 20 6 10 (1.2 6 0.6) 3 105
b 5 1.6, aCO 5 34 6 34{
Dust A 15.9 6 0.8{ 0.032 6 0.008 (2.8 6 0.7) 3 102
Gas1 A (36 6 9)dGD ((3.1 6 0.8) 3 105)dGD
H2 A 8.3 6 9 (0.5 6 0.5) 3 105
H2 " B 5.3 6 6 (0.3 6 0.3) 3 105
b 5 2, aCO 5 271 6 97{
Dust A 13.6 6 0.6{ 0.087 6 0.022 (7.5 6 1.9) 3 102
Gas1 A (95 6 24)dGD ((8.3 6 2.1) 3 105)dGD
H2 A 67 6 24 (3.9 6 1.4) 3 105
H2 " B 44 6 16 (2.5 6 0.9) 3 105
* Mass column density of gas or dust.
{ The units of aCO are M8 pc22 K21 km21 s. The uncertainty is dominated by the uncertainties in the 160-mm and 870-mm fluxes, as indicated by their error limits in Table 1. The error limits are approximately
symmetric.
{ The dust temperature in region A is assumed to be the same as the measured dust temperature in region A1.
1 dGD 5 RGD/1,100 is the gas-to-dust ratio, RGD, normalized by the solar value and scaled to the metallicity of WLM. Lowering dGD lowers aCO, but this does not seem reasonable: data suggests that the gas-to-dust
mass ratio is ,5,000 for 12 1 log(O/H) 5 7.8 (ref. 27), and this implies larger values of dGD (,4.5) and aCO. The gas mass and resulting aCO value also depend on the assumed correction factor of 1.7 for
submillimetre excess. With no correction for this excess, aCO increases for all b values: at b 5 1.8, aCO 5 370. Solutions with no submillimetre excess correction and lower b values19 can be found in Supplementary
Information. In addition, aCO depends on the assumed value of CO(3–2)/CO(1–0), which was taken to be 0.8 in Table 2; a value of CO(3–2)/CO(1–0) 5 1 increases aCO to 155 for b 5 1.8.
|| The H I mass column density is corrected for helium and heavy elements.
"The molecular mass for region B was calculated using the CO integrated intensity and the value of aCO determined from region A.
4 8 8 | N AT U R E | VO L 4 9 5 | 2 8 M A R C H 2 0 1 3
LETTER RESEARCH
combination of lower b with no submillimetre correction19 gives similar 2. Lee, H., Skillman, E. D. & Venn, K. A. Investigating the possible anomaly between
nebular and stellar oxygen abundances in the dwarf irregular galaxy WLM.
results (Supplementary Information). The dust mass was calculated Astrophys. J. 620, 223–237 (2005).
from an emissivity14 of k 5 13.9 cm2 gm21 at 140 mm, and converted 3. Asplund, M., Grevesse, N., Sauval, A. J. & Scott, P. The chemical composition of the
to 870 mm with the same power-law index, b. The dust mass for sun. Annu. Rev. Astron. Astrophys. 47, 481–522 (2009).
4. Leaman, R. et al. The resolved structure and dynamics of an isolated dwarf galaxy:
region A is then Mdust,A 5 S870,AD2/kBv(Td) for flux S870,A, distance D a VLT and Keck spectroscopic survey of WLM. Astrophys. J. 750, 33 (2012).
and black-body spectral function Bv. 5. Taylor, C. L. & Klein, U. A search for CO in the Local Group dwarf irregular galaxy
Dust mass is converted to gas mass using a factor equal to the gas-to- WLM. Astron. Astrophys. 366, 811–816 (2001).
6. Bigiel, F. et al. A constant molecular gas depletion time in nearby disk galaxies.
dust mass ratio, RGD. An approximation8 is to assume the solar value20 Astrophys. J. 730, L13 (2011).
(1/0.007) increased by the inverse of the metallicity of WLM (0.13), 7. Taylor, C. L., Kobulnicky, H. A. & Skillman, E. D. CO emission in low-luminosity,
which would give 1,100. We use this approximation here, but intro- H I-rich galaxies. Astron. J. 116, 2746–2756 (1998).
duce a scaling factor to the gas mass, dGD 5 RGD/1,100; that is, the gas- 8. Leroy, A. K. et al. The CO-to-H2 conversion factor from infrared dust emission
across the Local Group. Astrophys. J. 737, 12 (2011).
to-dust ratio normalized by the solar value and scaled to the metallicity. 9. Schruba, A. et al. Low CO luminosities in dwarf galaxies. Astron. J. 143, 138 (2012).
The results are in Table 2, assuming b 5 1.8 as the fiducial value and 10. Vassilev, V. et al. A Swedish heterodyne facility instrument for the APEX telescope.
comparing the results with those for b 5 1.6 and 2 to illustrate the Astron. Astrophys. 490, 1157–1163 (2008).
11. Siringo, G. et al. The Large APEX BOlometer CAmera LABOCA. Astron. Astrophys.
dependence of the results on b. Dust and gas mass correlate15 with the 497, 945–962 (2009).
assumed b value. The total gas mass column density in a 2299 region 12. Dale, D. A. et al. The Spitzer Local Volume Legacy: survey description and infrared
around region A is ,58 6 15M8 pc22 for b 5 1.8 (M8 , solar mass). photometry. Astrophys. J. 703, 517–556 (2009).
13. Planck Collaboration et al. Planck early results. XXV. Thermal dust in nearby
Atomic hydrogen contributes ,27.3 6 1.4M8 pc22, and the remain- molecular clouds. Astron. Astrophys. 536, A25 (2011).
der is ascribed to molecular H2 traced by the observed CO. 14. Draine, B. T. Interstellar dust grains. Annu. Rev. Astron. Astrophys. 41, 241–289 (2003).
The integral under the CO(3–2) line from region A is ICO 5 15. Galametz, M. et al. Mapping the cold dust temperatures and masses of nearby
KINGFISH galaxies with Herschel. Mon. Not. R. Astron. Soc. 425, 763–787 (2012).
0.200 6 0.046 K km s21. This intensity has to be converted to CO(1–0) 16. Coupeaud, et al. Low-temperature FIR and submillimetre mass absorption
before comparing it with H2 mass in the conventional way. We take as coefficient of interstellar silicate dust analogues. Astron. Astrophys. 535, A124 (2011).
a guide21 the value of CO(3–2)/CO(1–0) < 0.80 in another low- 17. Galametz, M. et al. Probing the dust properties of galaxies up to submillimetre
metallicity galaxy, the SMC22 (where O/H 5 20% of the solar value). wavelengths. II. Dust-to-gas mass ratio trends with metallicity and the submm
excess in dwarf galaxies. Astron. Astrophys. 532, A56 (2011).
The result, 0.25 K km s21, is combined with the H2 mass column 18. Verdugo, C. Sub-Millimeter Studies of Cold Dust and Gas in the Magellanic Clouds.
density to determine the conversion factor, aCO, from CO(1–0) to MSc thesis, Univ. Chile (2012).
H2. If aCO can be calibrated as a function of metallicity, then CO 19. Planck Collaboration. Planck early results. XVII. Origin of the submillimetre excess
dust emission in the Magellanic Clouds. Astron. Astrophys. 536, A17 (2011).
observations can be used directly to infer the molecular gas con- 20. Draine, B. T. et al. Dust masses, PAH abundances, and starlight intensities in the
tent irrespective of the dust spectral energy distribution. Extensive SINGS galaxy sample. Astrophys. J. 663, 866–894 (2007).
compilations7–9 show aCO increasing strongly at lower metallicity, from 21. Nikolić, S., Garay, G., Rubio, M. & Johansson, L. E. B. CO and CS in the Magellanic
Clouds: a x2-analysis of multitransitional data based on the MEP radiative transfer
,4M8 pc22 K21 km21 s in the Milky Way8, where the metallicity3 is model. Astron. Astrophys. 471, 561–571 (2007).
12 1 log(O/H) 5 8.69, down to the previous CO detection limit8 in the 22. Dufour, R. J. The composition of H II regions in the Magellanic Clouds. IAU Symp.
SMC, where aCO < 70M8 pc22 K21 km21 s at 12 1 log(O/H) 5 8.0. 108, 353–360 (1984).
23. Hunter, D. A., Elmegreen, B. G. & Ludka, B. C. GALEX ultraviolet imaging of dwarf
Our observations of WLM2 at a metallicity of 12 1 log(O/H) 5 7.8 galaxies and star formation rates. Astron. J. 139, 447–475 (2010).
continue this trend. 24. Leroy, A. K. et al. The star formation efficiency in nearby galaxies: measuring where
Taking the H2 column density from the residual between the dust- gas forms stars effectively. Astron. J. 136, 2782–2845 (2008).
derived total and the H I column density, 31 6 15M8 pc22, and divid- 25. Lada, C. J., Forbrich, J., Lombardi, M. & Alves, J. F. Star formation rates in molecular clouds
and the nature of the extragalactic scaling relations. Astrophys. J. 745, 190 (2012).
ing by the inferred CO(1–0) line integral of 0.25 K km s21, we obtain 26. Mannucci, F. et al. LSD: Lyman-break galaxies, stellar populations and dynamics – I.
aCO 5 124 6 60M8 pc22 K21 km21 s including helium and heavy Mass, metallicity and gas at z , 3.1. Mon. Not. R. Astron. Soc. 398, 1915–1931 (2009).
elements, with a range in aCO from 34 to 271 as b varies from 1.6 27. Engelbracht, C. W. et al. Metallicity effects on dust properties in starbursting
galaxies. Astrophys. J. 678, 804–827 (2008).
to 2. The corresponding factor, XCO, for conversion from ICO to H2 28. Zhang, H.-X., Hunter, D. A., Elmegreen, B. G., Gao, Y. & Schruba, A. Outside-in shrinking
column density would be (5.8 6 2.8) 3 1021 cm22 K21 km21 s, ranging of the star-forming disks of dwarf irregular galaxies. Astron. J. 143, 47 (2012).
from 1.5 3 1021 to 1.3 3 1022 as b varies between 1.6 and 2. There is a 29. McMillan, P. J. Mass models of the Milky Way. Mon. Not. R. Astron. Soc. 414,
2446–2457 (2011).
large uncertainty because of the unknown dust properties (b, k, dGD 30. Chomiuk, L. & Povich, M. S. Toward a unification of star formation rate
and the submillimetre excess) and molecular excitation (CO(3–2)/ determinations in the Milky Way and other galaxies. Astron. J. 142, 197 (2011).
CO(1–0)) in dIrr galaxies. Supplementary Information is available in the online version of the paper.
The star-formation rate based on the Ha and far-ultraviolet23 fluxes
within an 1899 aperture centred on cloud A is (3.9–4.8) 3 1025M8 yr21. Acknowledgements This work was funded in part by the US National Science
Foundation through grants AST-0707563 and AST-0707426 to D.A.H. and B.G.E. M.R.
Dividing these rates into the CO-associated molecular mass using and C.V. wish to acknowledge support from CONICYT (FONDECYT grant no. 1080335).
aCO 5 124 M8 pc22 K21 km21 s gives a CO molecular consumption M.R. was also supported by the Chilean Center for Astrophysics FONDAP grant
time (for converting gas into stars) of 4.6–3.8 Gyr for region A. In no. 15010003. A.S. was supported by the Deutsche Forschungsgemeinschaft Priority
Program 1177. We are grateful to M. Albrecht for help with the LABOCA data reduction
region B, the star-formation rate from Ha and far-ultraviolet fluxes and to L. Hill for making Fig. 1a. The National Radio Astronomy Observatory is a facility
is (1.7–12.6) 3 1025M8 yr21 and the CO molecular consumption of the US National Science Foundation operated under cooperative agreement by
time is 6.7–1.5 Gyr. These times are only slightly larger than the aver- Associated Universities, Inc.
age value in spiral galaxies24, which is ,2 Gyr, but they are ten times Author Contributions B.G.E. coordinated the observational team, did the calculations
larger than the rate per molecule in local giant molecular clouds25, for Table 2 and wrote the manuscript; M.R. was principal investigator for Chilean
observing time on the APEX telescope and, with C.V., observed the galaxy in CO and at
which is a more direct analogy with our observations. 870 mm, reduced the relevant data in Table 1 and did relevant calculations for Table 2;
The detection of CO in WLM suggests that star formation continues D.A.H. determined the observational strategy, selected WLM for study, chose the
to occur in dense molecular gas even at lower metallicities than prev- observing coordinates, extracted the H I spectra from the LITTLE THINGS data and
prepared Fig. 1. E.B. was principal investigator on the APEX proposal using European
iously observed. The similarity between the metallicities of dIrr galaxies time through ESO and coordinated the work on data uncertainties and background
such as WLM and those of larger galaxies at high redshift26 implies that noise. A.S. made the WLM H I data cube from Jansky Very Large Array observations. All
we should be able to study star formation in young galaxies using the authors discussed the results and commented on the manuscript.
usual techniques. Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Received 24 October 2012; accepted 23 January 2013. Readers are welcome to comment on the online version of the paper. Correspondence
and requests for materials should be addressed to and requests for materials should be
1. Krumholz, M. R. Star formation in atomic gas. Astrophys. J. 759, 9 (2012). addressed to B.G.E. ([email protected]).
2 8 M A R C H 2 0 1 3 | VO L 4 9 5 | N AT U R E | 4 8 9
Bibliography
Alberts, S., Calzetti, D., Dong, H. et al. The Evolution of Stellar Populations
in the Outer Disks of Spiral Galaxies. ApJ, 2011. vol. 731, 28.
Allen, R. J., Atherton, P. D. & Tilanus, R. P. J. Large-scale dissociation of
molecular gas in galaxies by newly formed stars. Nature, 1986. vol. 319, pp.
296–298.
Allen, R. J., Heaton, H. I. & Kaufman, M. J. The Production of H I in Pho-
todissociation Regions and a Comparison with CO (1-0) Emission. ApJ, 2004.
vol. 608, pp. 314–322.
Battaglia, G., Fraternali, F., Oosterloo, T. et al. ion{H}{i} study of the
warped spiral galaxy NGC 5055: a disk/dark matter halo offset? . A&A, 2006. vol.
447, pp. 49–62.
Bauermeister, A., Blitz, L. & Ma, C.-P. The Gas Consumption History to
Redshift 4 . ApJ, 2010. vol. 717, pp. 323–332.
Bendo, G. J., Galliano, F. & Madden, S. C. MIPS 24-160 µm photometry for
the Herschel-SPIRE Local Galaxies Guaranteed Time Programs. MNRAS, 2012.
vol. 423, pp. 197–212.
Bigiel, F., Leroy, A., Walter, F. et al. The Star Formation Law in Nearby
Galaxies on Sub-Kpc Scales. AJ, 2008. vol. 136, pp. 2846–2871.
Bigiel, F., Leroy, A., Walter, F. et al. Extremely Inefficient Star Formation in
the Outer Disks of Nearby Galaxies. AJ, 2010. vol. 140, pp. 1194–1213.
Bigiel, F., Leroy, A. K., Walter, F. et al. A Constant Molecular Gas Depletion
Time in Nearby Disk Galaxies. ApJ, 2011. vol. 730, L13.
Boissier, S., Boselli, A., Duc, P.-A. et al. The GALEX Ultraviolet Virgo Cluster
Survey (GUViCS). II. Constraints on star formation in ram-pressure stripped gas.
A&A, 2012. vol. 545, A142.
Boissier, S., Gil de Paz, A., Boselli, A. et al. GALEX Observations of Low
Surface Brightness Galaxies: UV Color and Star Formation Efficiency. ApJ,
2008. vol. 681, pp. 244–257.
Bolatto, A. D., Wolfire, M. & Leroy, A. K. The CO-to-H2 Conversion Factor .
ARA&A, 2013. vol. 51, pp. 207–268.
116
Bouché, N., Murphy, M. T., Kacprzak, G. G. et al. Signatures of Cool Gas
Fueling a Star-Forming Galaxy at Redshift 2.3 . Science, 2013. vol. 341, pp. 50–53.
Braine, J., Ferguson, A. M. N., Bertoldi, F. et al. The Detection of Molecular
Gas in the Outskirts of NGC 6946 . ApJ, 2007. vol. 669, pp. L73–L76.
Braine, J., Gratier, P., Kramer, C. et al. Molecular cloud formation and the
star formation efficiency in M 33. Molecule and star formation in M 33 . A&A,
2010. vol. 520, A107.
Braine, J. & Herpin, F. Molecular hydrogen beyond the optical edge of an isolated
spiral galaxy. Nature, 2004. vol. 432, pp. 369–371.
Braine, J., Lisenfeld, U., Due, P.-A. et al. Formation of molecular gas in the
tidal debris of violent galaxy-galaxy interactions. Nature, 2000. vol. 403, pp.
867–869.
Carter, M., Lazareff, B., Maier, D. et al. The EMIR multi-band mm-wave
receiver for the IRAM 30-m telescope. A&A, 2012. vol. 538, A89.
Casasola, V., Hunt, L., Combes, F. et al. The resolved star-formation relation
in nearby active galactic nuclei . ArXiv e-prints, 2015.
Cayatte, V., van Gorkom, J. H., Balkowski, C. et al. VLA observations of
neutral hydrogen in Virgo Cluster galaxies. I - The Atlas. AJ, 1990. vol. 100, pp.
604–634.
Chung, A., van Gorkom, J. H., Kenney, J. D. P. et al. Virgo Galaxies with
Long One-sided H I Tails. ApJ, 2007. vol. 659, pp. L115–L119.
Chung, A., van Gorkom, J. H., Kenney, J. D. P. et al. VLA Imaging of Virgo
Spirals in Atomic Gas (VIVA). I. The Atlas and the H I Properties. AJ, 2009.
vol. 138, pp. 1741–1816.
Cluver, M. E., Appleton, P. N., Boulanger, F. et al. Powerful H2 Line Cooling
in Stephan’s Quintet. I. Mapping the Significant Cooling Pathways in Group-wide
Shocks. ApJ, 2010. vol. 710, pp. 248–264.
Combes, F. Gas Accretion in Disk Galaxies. M. S. Seigar & P. Treuthardt, eds.,
Structure and Dynamics of Disk Galaxies, vol. 480 of Astronomical Society of the
Pacific Conference Series. p. 211.
Conselice, C. J., Gallagher, J. S., III & Wyse, R. F. G. On the Nature of the
NGC 1275 System. AJ, 2001. vol. 122, pp. 2281–2300.
Cortese, L., Marcillac, D., Richard, J. et al. The strong transformation of
spiral galaxies infalling into massive clusters at z ˜ 0.2 . MNRAS, 2007. vol. 376,
pp. 157–172.
Coupeaud, A., Demyk, K., Meny, C. et al. Low-temperature FIR and submil-
limetre mass absorption coefficient of interstellar silicate dust analogues. A&A,
2011. vol. 535, A124.
117
Crosthwaite, L. P., Turner, J. L., Buchholz, L. et al. CO in the Disk of the
Barred Spiral Galaxy M83: CO (1-0), CO (2-1), and Neutral Gas. AJ, 2002. vol.
123, pp. 1892–1912.
Dale, D. A., Cohen, S. A., Johnson, L. C. et al. The Spitzer Local Volume
Legacy: Survey Description and Infrared Photometry. ApJ, 2009. vol. 703, pp.
517–556.
Dasyra, K. M., Combes, F., Salomé, P. et al. Survival of molecular gas in Virgo’s
hot intracluster medium: CO near M 86 . A&A, 2012. vol. 540, A112.
Davies, J. I., Bianchi, S., Cortese, L. et al. The Herschel Virgo Cluster Survey
- VIII. The Bright Galaxy Sample. MNRAS, 2012. vol. 419, pp. 3505–3520.
DeMaio, T., Gonzalez, A. H., Zabludoff, A. et al. On the origin of the intra-
cluster light in massive galaxy clusters. MNRAS, 2015. vol. 448, pp. 1162–1177.
Dessauges-Zavadsky, M., Verdugo, C., Combes, F. et al. CO map and steep
Kennicutt-Schmidt relation in the extended UV disk of M 63 . A&A, 2014. vol.
566, A147.
Dickman, R. L., Snell, R. L. & Schloerb, F. P. Carbon monoxide as an extra-
galactic mass tracer . ApJ, 1986. vol. 309, pp. 326–330.
Dobbs, C. L., Krumholz, M. R., Ballesteros-Paredes, J. et al. Formation
of Molecular Clouds and Global Conditions for Star Formation. Protostars and
Planets VI, 2014, pp. 3–26.
Dong, H., Calzetti, D., Regan, M. et al. Spitzer Observations of Star Formation
in the Extreme Outer Disk of M83 (NGC5236). AJ, 2008. vol. 136, 479.
Draine, B. T. Interstellar Dust Grains. ARA&A, 2003. vol. 41, pp. 241–289.
Draine, B. T., Dale, D. A., Bendo, G. et al. Dust Masses, PAH Abundances,
and Starlight Intensities in the SINGS Galaxy Sample. ApJ, 2007. vol. 663, pp.
866–894.
Drazinos, P., Kontizas, E., Bellas-Velidis, I. et al. Indication of Hierarchical
Star Formation in Spiral Galaxies. 2013. pp. 36–47.
Dufour, R. J. The composition of H II regions in the Magellanic Clouds. S. van
den Bergh & K. S. D. de Boer, eds., Structure and Evolution of the Magellanic
Clouds, vol. 108 of IAU Symposium. pp. 353–360.
Edge, A. C., Oonk, J. B. R., Mittal, R. et al. Herschel observations of FIR
emission lines in brightest cluster galaxies. A&A, 2010. vol. 518, L46.
Elmegreen, D. M. & Elmegreen, B. Hierarchical Star Formation in LEGUS
Galaxies. American Astronomical Society Meeting Abstracts 224 , vol. 224 of
American Astronomical Society Meeting Abstracts. p. 223.
Fabian, A. C., Sanders, J. S., Taylor, G. B. et al. A very deep Chandra obser-
vation of the Perseus cluster: shocks, ripples and conduction. MNRAS, 2006. vol.
366, pp. 417–428.
118
Feldmeier, J. J., Mihos, J. C., Morrison, H. L. et al. Deep CCD Surface
Photometry of Galaxy Clusters. I. Methods and Initial Studies of Intracluster
Starlight. ApJ, 2002. vol. 575, pp. 779–800.
Fouqué, P., Solanes, J. M., Sanchis, T. et al. Structure, mass and distance of
the Virgo cluster from a Tolman-Bondi model . A&A, 2001. vol. 375, pp. 770–780.
Galametz, M., Kennicutt, R. C., Albrecht, M. et al. Mapping the cold dust
temperatures and masses of nearby KINGFISH galaxies with Herschel . MNRAS,
2012. vol. 425, pp. 763–787.
Galametz, M., Madden, S. C., Galliano, F. et al. Probing the dust properties
of galaxies up to submillimetre wavelengths. II. Dust-to-gas mass ratio trends with
metallicity and the submm excess in dwarf galaxies. A&A, 2011. vol. 532, A56.
Gao, Y. & Solomon, P. M. The Star Formation Rate and Dense Molecular Gas
in Galaxies. ApJ, 2004. vol. 606, pp. 271–290.
Gavazzi, G., Boselli, A., Mayer, L. et al. 75 Kiloparsec Trails of Ionized Gas
behind Two Irregular Galaxies in A1367 . ApJ, 2001. vol. 563, pp. L23–L26.
Gil de Paz, A., Boissier, S., Madore, B. F. et al. The GALEX Ultraviolet Atlas
of Nearby Galaxies. ApJS, 2007a. vol. 173, pp. 185–255.
Gil de Paz, A., Madore, B. F., Boissier, S. et al. Discovery of an Extended
Ultraviolet Disk in the Nearby Galaxy NGC 4625 . ApJ, 2005. vol. 627, pp. L29–
L32.
Gil de Paz, A., Madore, B. F., Boissier, S. et al. Chemical and Photometric
Evolution of Extended Ultraviolet Disks: Optical Spectroscopy of M83 (NGC 5236)
and NGC 4625 . ApJ, 2007b. vol. 661, pp. 115–134.
Gong, M. & Ostriker, E. C. Prestellar Core Formation, Evolution, and Accretion
from Gravitational Fragmentation in Turbulent Converging Flows. ApJ, 2015. vol.
806, 31.
Gu, L., Yagi, M., Nakazawa, K. et al. Multi-wavelength Studies of Spectacular
Ram Pressure Stripping of a Galaxy: Discovery of an X-Ray Absorption Feature.
ApJ, 2013. vol. 777, L36.
Gunn, J. E. & Gott, J. R., III. On the Infall of Matter Into Clusters of Galaxies
and Some Effects on Their Evolution. ApJ, 1972. vol. 176, p. 1.
Haynes, M. P. & Giovanelli, R. Neutral hydrogen in isolated galaxies. IV - Results
for the Arecibo sample. AJ, 1984. vol. 89, pp. 758–800.
Helfer, T. T., Thornley, M. D., Regan, M. W. et al. The BIMA Survey of
Nearby Galaxies (BIMA SONG). II. The CO Data. ApJS, 2003. vol. 145, pp.
259–327.
Henry, R. B. C. & Worthey, G. The Distribution of Heavy Elements in Spiral
and Elliptical Galaxies. PASP, 1999. vol. 111, pp. 919–945.
119
Holwerda, B. W., Pirzkal, N. & Heiner, J. S. Quantified H I morphology - VI.
The morphology of extended discs in UV and H I . MNRAS, 2012. vol. 427, pp.
3159–3175.
Hunter, D. A., Elmegreen, B. G. & Ludka, B. C. Galex Ultraviolet Imaging of
Dwarf Galaxies and Star Formation Rates. AJ, 2010. vol. 139, pp. 447–475.
Isobe, T., Feigelson, E. D., Akritas, M. G. et al. Linear regression in astron-
omy.. ApJ, 1990. vol. 364, pp. 104–113.
Iwasawa, K., Wilson, A. S., Fabian, A. C. et al. The X-ray nebula around the
type 2 Seyfert galaxy NGC 4388 . MNRAS, 2003. vol. 345, pp. 369–378.
Jáchym, P., Combes, F., Cortese, L. et al. Abundant Molecular Gas and Ineffi-
cient Star Formation in Intracluster Regions: Ram Pressure Stripped Tail of the
Norma Galaxy ESO137-001 . ApJ, 2014. vol. 792, 11.
Jáchym, P., Palouš, J., Köppen, J. et al. Gas stripping in galaxy clusters: a new
SPH simulation approach. A&A, 2007. vol. 472, pp. 5–20.
Kenney, J. D. P., Tal, T., Crowl, H. H. et al. A Spectacular Hα Complex in
Virgo: Evidence for a Collision between M86 and NGC 4438 and Implications for
the Collisional ISM Heating of Ellipticals. ApJ, 2008. vol. 687, pp. L69–L74.
Kenney, J. D. P., van Gorkom, J. H. & Vollmer, B. VLA H I Observations
of Gas Stripping in the Virgo Cluster Spiral NGC 4522 . AJ, 2004. vol. 127, pp.
3361–3374.
Kenney, J. D. P. & Young, J. S. The effects of environment on the molecular
and atomic gas properties of large Virgo cluster spirals. ApJ, 1989. vol. 344, pp.
171–199.
Kennicutt, R. C., Calzetti, D., Aniano, G. et al. KINGFISH – Key Insights
on Nearby Galaxies: A Far-Infrared Survey with Herschel: Survey Description
and Image Atlas. PASP, 2011. vol. 123, pp. 1347–1369.
Kennicutt, R. C. & Evans, N. J. Star Formation in the Milky Way and Nearby
Galaxies. ARA&A, 2012. vol. 50, pp. 531–608.
Kennicutt, R. C., Jr. The star formation law in galactic disks. ApJ, 1989. vol.
344, pp. 685–703.
Kennicutt, R. C., Jr. Star Formation in Galaxies Along the Hubble Sequence.
ARA&A, 1998a. vol. 36, pp. 189–232.
Kennicutt, R. C., Jr. The Global Schmidt Law in Star-forming Galaxies. ApJ,
1998b. vol. 498, pp. 541–552.
Kennicutt, R. C., Jr., Lee, J. C., Funes, S. J., José G. et al. An Hα Imaging
Survey of Galaxies in the Local 11 Mpc Volume. ApJS, 2008. vol. 178, pp. 247–279.
Kroupa, P. & Weidner, C. Galactic-Field Initial Mass Functions of Massive
Stars. ApJ, 2003. vol. 598, pp. 1076–1078.
120
Lada, C. J., Forbrich, J., Lombardi, M. et al. Star Formation Rates in Molecular
Clouds and the Nature of the Extragalactic Scaling Relations. ApJ, 2012. vol. 745,
190.
Leroy, A. K., Bolatto, A., Gordon, K. et al. The CO-to-H2 Conversion Factor
from Infrared Dust Emission across the Local Group. ApJ, 2011. vol. 737, 12.
Leroy, A. K., Walter, F., Bigiel, F. et al. Heracles: The HERA CO Line
Extragalactic Survey. AJ, 2009. vol. 137, pp. 4670–4696.
Leroy, A. K., Walter, F., Brinks, E. et al. The Star Formation Efficiency in
Nearby Galaxies: Measuring Where Gas Forms Stars Effectively. AJ, 2008. vol.
136, pp. 2782–2845.
Leroy, A. K., Walter, F., Sandstrom, K. et al. Molecular Gas and Star For-
mation in nearby Disk Galaxies. AJ, 2013. vol. 146, 19.
Lim, J., Ohyama, Y., Chi-Hung, Y. et al. A Molecular Hydrogen Nebula in the
Central cD Galaxy of the Perseus Cluster . ApJ, 2012. vol. 744, 112.
Longmore, S. N., Bally, J., Testi, L. et al. Variations in the Galactic star
formation rate and density thresholds for star formation. MNRAS, 2013. vol.
429, pp. 987–1000.
Lupton, R., Blanton, M. R., Fekete, G. et al. Preparing Red-Green-Blue Images
from CCD Data. PASP, 2004. vol. 116, pp. 133–137.
Machacek, M., Dosaj, A., Forman, W. et al. Infall of the Elliptical Galaxy NGC
1404 into the Fornax Cluster . ApJ, 2005. vol. 621, pp. 663–672.
Machacek, M. E., Jones, C. & Forman, W. R. Chandra Observations of NGC
4438: An Environmentally Damaged Galaxy in the Virgo Cluster . ApJ, 2004. vol.
610, pp. 183–200.
Mahdavi, A., Geller, M. J., Fabricant, D. G. et al. The Lumpy Cluster Abell
1185 . AJ, 1996. vol. 111, p. 64.
Mannucci, F., Cresci, G., Maiolino, R. et al. LSD: Lyman-break galaxies Stellar
populations and Dynamics - I. Mass, metallicity and gas at z ˜ 3.1 . MNRAS, 2009.
vol. 398, pp. 1915–1931.
Martin, C. L. & Kennicutt, R. C., Jr. Star Formation Thresholds in Galactic
Disks. ApJ, 2001. vol. 555, pp. 301–321.
Mei, S., Blakeslee, J. P., Côté, P. et al. The ACS Virgo Cluster Survey. XIII.
SBF Distance Catalog and the Three-dimensional Structure of the Virgo Cluster .
ApJ, 2007. vol. 655, pp. 144–162.
Merritt, D. Relaxation and tidal stripping in rich clusters of galaxies. II. Evolution
of the luminosity distribution. ApJ, 1984. vol. 276, pp. 26–37.
Mihos, J. C., Harding, P., Feldmeier, J. et al. Diffuse Light in the Virgo Cluster .
ApJ, 2005. vol. 631, pp. L41–L44.
121
Narayanan, D., Krumholz, M., Ostriker, E. C. et al. The CO-H2 conversion
factor in disc galaxies and mergers. MNRAS, 2011. vol. 418, pp. 664–679.
Nieten, C., Neininger, N., Guélin, M. et al. Molecular gas in the Andromeda
galaxy. A&A, 2006. vol. 453, pp. 459–475.
Nikolić, S., Garay, G., Rubio, M. et al. CO and CS in the Magellanic Clouds: a
χˆ2-analysis of multitransitional data based on the MEP radiative transfer model .
A&A, 2007. vol. 471, pp. 561–571.
Nulsen, P. E. J. Transport processes and the stripping of cluster galaxies. MNRAS,
1982. vol. 198, pp. 1007–1016.
Oosterloo, T. & van Gorkom, J. A large H I cloud near the centre of the Virgo
cluster . A&A, 2005. vol. 437, pp. L19–L22.
O’Sullivan, E., Giacintucci, S., Vrtilek, J. M. et al. A Chandra X-ray View
of Stephan’s Quintet: Shocks and Star Formation. ApJ, 2009. vol. 701, pp. 1560–
1568.
Paturel, G., Petit, C., Prugniel, P. et al. HYPERLEDA. I. Identification and
designation of galaxies. A&A, 2003. vol. 412, pp. 45–55.
Planck Collaboration, Abergel, A., Ade, P. A. R. et al. Planck early results.
XXV. Thermal dust in nearby molecular clouds. A&A, 2011. vol. 536, A25.
Pohlen, M. & Trujillo, I. The structure of galactic disks. Studying late-type
spiral galaxies using SDSS . A&A, 2006. vol. 454, pp. 759–772.
Quilis, V., Moore, B. & Bower, R. Gone with the Wind: The Origin of S0
Galaxies in Clusters. Science, 2000. vol. 288, pp. 1617–1620.
Roediger, E. & Hensler, G. Ram pressure stripping of disk galaxies. From high
to low density environments. A&A, 2005. vol. 433, pp. 875–895.
Salomé, P., Combes, F., Edge, A. C. et al. Cold molecular gas in the Perseus
cluster core. Association with X-ray cavity, Hα filaments and cooling flow . A&A,
2006. vol. 454, pp. 437–445.
Salomé, P., Combes, F., Revaz, Y. et al. A very extended molecular web around
NGC 1275 . A&A, 2011. vol. 531, A85.
Sánchez Almeida, J., Elmegreen, B. G., Muñoz-Tuñón, C. et al. Star for-
mation sustained by gas accretion. A&A Rev., 2014. vol. 22, 71.
Sancisi, R., Fraternali, F., Oosterloo, T. et al. Cold gas accretion in galaxies.
A&A Rev., 2008. vol. 15, pp. 189–223.
Sandstrom, K. M., Leroy, A. K., Walter, F. et al. The CO-to-H2 Conversion
Factor and Dust-to-gas Ratio on Kiloparsec Scales in Nearby Galaxies. ApJ, 2013.
vol. 777, 5.
Schmidt, M. The Rate of Star Formation.. ApJ, 1959. vol. 129, p. 243.
122
Schruba, A., Leroy, A. K., Walter, F. et al. A Molecular Star Formation Law
in the Atomic-gas-dominated Regime in Nearby Galaxies. AJ, 2011. vol. 142, 37.
Scott, T. C., Cortese, L., Brinks, E. et al. Two long H I tails in the outskirts
of Abell 1367 . MNRAS, 2012. vol. 419, pp. L19–L23.
Serra, P., Koribalski, B., Duc, P.-A. et al. Discovery of a giant HI tail in the
galaxy group HCG 44 . MNRAS, 2013. vol. 428, pp. 370–380.
Shlosman, I., Frank, J. & Begelman, M. C. Bars within bars - A mechanism
for fuelling active galactic nuclei . Nature, 1989. vol. 338, pp. 45–47.
Smith, D. A., Allen, R. J., Bohlin, R. C. et al. A New Probe of the Molecular
Gas in Galaxies: Application to M101 . ApJ, 2000. vol. 538, pp. 608–622.
Solomon, P. M., Rivolo, A. R., Barrett, J. et al. Mass, luminosity, and line
width relations of Galactic molecular clouds. ApJ, 1987. vol. 319, pp. 730–741.
Spitzer, L., Jr. & Baade, W. Stellar Populations and Collisions of Galaxies..
ApJ, 1951. vol. 113, p. 413.
Sun, M., Donahue, M., Roediger, E. et al. Spectacular X-ray Tails, Intracluster
Star Formation, and ULXs in A3627 . ApJ, 2010. vol. 708, pp. 946–964.
Sun, M., Donahue, M. & Voit, G. M. Hα Tail, Intracluster H II Regions, and
Star Formation: ESO 137-001 in Abell 3627 . ApJ, 2007. vol. 671, pp. 190–202.
Tamura, T., Maeda, Y., Mitsuda, K. et al. X-ray Spectroscopy of the Core of the
Perseus Cluster with Suzaku: Elemental Abundances in the Intracluster Medium.
ApJ, 2009. vol. 705, pp. L62–L66.
Thilker, D. A., Bianchi, L., Boissier, S. et al. Recent Star Formation in the
Extreme Outer Disk of M83 . ApJ, 2005. vol. 619, pp. L79–L82.
Tilanus, R. P. J. & Allen, R. J. Spiral Structure of M83 - Distribution and
Kinematics of the Atomic and Ionized Hydrogen. A&A, 1993. vol. 274, p. 707.
Tonnesen, S., Bryan, G. L. & van Gorkom, J. H. Environmentally Driven
Evolution of Simulated Cluster Galaxies. ApJ, 2007. vol. 671, pp. 1434–1445.
Verdes-Montenegro, L., Bosma, A. & Athanassoula, E. A detailed study of
the ringed galaxy NGC 3344 . A&A, 2000. vol. 356, pp. 827–839.
Verdugo, Celia. Master’s thesis: Sub-Millimeter Studies of Cold Dust and Gas in
the Magellanic Clouds. MSc thesis, Univ. Chile, 2012.
Vollmer, B., Braine, J., Combes, F. et al. New CO observations and simulations
of the NGC 4438/NGC 4435 system. Interaction diagnostics of the Virgo cluster
galaxy NGC 4438 . A&A, 2005. vol. 441, pp. 473–489.
Vollmer, B., Braine, J., Pappalardo, C. et al. Ram-pressure stripped molecular
gas in the Virgo spiral galaxy NGC 4522 . A&A, 2008. vol. 491, pp. 455–464.
Vollmer, B., Cayatte, V., Balkowski, C. et al. Ram Pressure Stripping and
Galaxy Orbits: The Case of the Virgo Cluster . ApJ, 2001. vol. 561, pp. 708–726.
123
Walter, F., Brinks, E., de Blok, W. J. G. et al. THINGS: The H I Nearby
Galaxy Survey. AJ, 2008. vol. 136, 2563.
Watson, L. C., Martini, P., Lisenfeld, U. et al. The Molecular Gas - Star
Formation Connection in an Extended Ultraviolet (XUV) Disk . American Astro-
nomical Society Meeting Abstracts 223 , vol. 223 of American Astronomical Society
Meeting Abstracts. p. 454.22.
Wu, J., Evans, N. J., II, Gao, Y. et al. Connecting Dense Gas Tracers of Star
Formation in our Galaxy to High-z Star Formation. ApJ, 2005. vol. 635, pp.
L173–L176.
Wyder, T. K., Martin, D. C., Barlow, T. A. et al. The Star Formation Law at
Low Surface Density. ApJ, 2009. vol. 696, pp. 1834–1853.
Yagi, M., Gu, L., Fujita, Y. et al. Multi-wavelength Studies of Spectacular Ram-
pressure Stripping of a Galaxy. II. Star Formation in the Tail . ApJ, 2013. vol.
778, 91.
Yagi, M., Komiyama, Y., Yoshida, M. et al. The Remarkable 60×2 kpc Optical
Filament Associated with a Poststarburst Galaxy in the Coma Cluster . ApJ, 2007.
vol. 660, pp. 1209–1214.
Yoshida, M., Yagi, M., Okamura, S. et al. Discovery of a Very Extended
Emission-Line Region around the Seyfert 2 Galaxy NGC 4388 . ApJ, 2002. vol.
567, pp. 118–129.
Young, J. S. & Scoville, N. Z. Molecular gas in galaxies. ARA&A, 1991. vol. 29,
pp. 581–625.
Young, J. S., Xie, S., Tacconi, L. et al. The FCRAO Extragalactic CO Survey.
I. The Data. ApJS, 1995. vol. 98, p. 219.
Zhang, B., Sun, M., Ji, L. et al. The Narrow X-Ray Tail and Double Hα Tails of
ESO 137-002 in A3627 . ApJ, 2013. vol. 777, 122.
Zhang, Q. Fragmentation of Molecular Clumps and Formation of Massive Cores.
Dense Cores: Origin, Evolution, and Collapse. p. 30201.
Zhang, Q., Wang, K., Lu, X. et al. Fragmentation of Molecular Clumps and
Formation of a Protocluster . ApJ, 2015. vol. 804, 141.
124

Thesis

Uploaded by

Copyright:

Available Formats

Thesis

Uploaded by

Document Information

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

Thesis

Uploaded by

Copyright:

Available Formats

OBSERVATOIRE DE PARIS

AUTHOR: CELIA ANAHI VERDUGO SALGADO

STAR FORMATION IN LOW GAS DENSITY AND LOW

Thesis supervised by FRANÇOISE COMBES

Defended on July 21st 2015, in Paris

AUTEUR: CELIA ANAHI VERDUGO SALGADO

FORMATION DES ÉTOILES DANS LES

Thèse dirigée par FRANÇOISE COMBES

Soutenue à Paris le 21 Juillet 2015

4 Summary and Conclusions 102

1.1 H2 energy levels diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.1 NUV-FUV color image of M63. . . . . . . . . . . . . . . . . . . . . . 26

1 Color composite image of the WLM galaxy. . . . . . . . . . . . . . . 109

2.1 CO mapping along the major axis of the M63 disk. . . . . . . . . . . 30

3.1 Targets and Observations. . . . . . . . . . . . . . . . . . . . . . . . . 82

Figure 1.1: H2 energy levels diagram

1.2 Molecular gas mass

NH2 = XCO ICO(1−0) (1.2)

MH2 = XCO µmH2 πR2 ICO(1−0) (1.3)

MH2 [M ] = 4.4πR2 [pc]ICO(1−0) [K km/s] (1.4)

where R is the radius of the CO(1-0) beamsize in parsecs, so it depends on the

1.3 Kennicutt-Schmidt relation

1.4 Low gas density environments

30s 13h 16m 00s 30s 13h 15m 00s

CO(2–1) IRAM 13.4 Leroy et al. (2009)

0 100 200 300 400 500 600 700

Star formation rate, efficiency, and Kennicutt-Schmidt relation

Our acquired CO measurements offer the opportunity to analyse the Kennicutt-

which includes a factor of 1.36 to reflect the presence of helium.

ΣH2 (M pc−2 ) = 4.4 cos(i)FCO(1−0) (K km s−1 ) (2.3)

0 100 200 300 400 500 600

log(ΣSFR) [M⊙ yr−1 kpc−2]

H2 -1.93 1.06 0.29 0.17† 3.0† ··· -2.22 0.92 0.10

2.1.4 Summary and conclusions

Deep CO(1–0) and CO(2–1) observations obtained on the IRAM 30 m telescope

CO map and steep Kennicutt-Schmidt relation in the extended

1. Introduction galactocentric radii. Recent star formation within such environ-

molecular clouds. Second, this suggests the presence of large

Table 1. CO mapping along the major axis of the M 63 disk.

Table 2. CO mapping over the bright UV region at 1.36 r25 .

Table 3. Public ancillary data of M 63.

Band or line Telescope FWHM Reference

FCO(2−1) < 0.06 K km s−1 , obtained when summing up all the

0 100 200 300 400 500 600 700

log SFE [yr−1] = log ΣSFR/ΣH2

log ΣH2 /ΣH I

This work Bigiel et al. 2008

log(ΣSFR) [M yr−1 kpc−2]

2.2.1 Data reduction and Results

Table 2.7. Galaxies selected for XUV observations in 2013.

NGC628 01:36:41.8 15:47:00.5 11 5.0±0.1 ···

a Information taken from Gil de Paz et al. (2007a).

Note. — Due to bad weather conditions NGC628 could not be observed.

NGC2403-1 07:35:31.42 65:38:40.78 0.4 no

Table 2.9. Line parameters for detections in XUV observations 2013.

NGC2403 CO(1–0) 4 126.0±1.5 40.3±3.5 60 10 2.58±0.19 1.04

NGC2403-1 3.4 <0.31 <1.25

Note. — Rms values correspond to Tmb scale,

molecular gas (e.g. Crosthwaite et al. 2002, Nieten et al. 2006).

2.3.1 Instrumentation and Observations