A 500 pc filamentary gas wisp in the disk of the Milky Way

Abstract

Star formation occurs in molecular gas. In previous studies, the structure of
the molecular gas has been studied in terms of molecular clouds, but has been
overlooked beyond the cloud scale. We present an observational study of the
molecular gas at 49.5 degree <l<52.5 degree and -5.0 km/s <v_lsr <17.4 km/s.
The molecular gas is found in the form of a huge (>= 500 pc) filamentary gas
wisp. This has a large physical extent and a velocity dispersion of ~5 km/s.
The eastern part of the filamentary gas wisp is located ~130 pc above the
Galactic disk (which corresponds to 1.5-4 e-folding scale-heights), and the
total mass of the gas wisp is >= 1 X 10^5 M_sun. It is composed of two
molecular clouds and an expanding bubble. The velocity structure of the gas
wisp can be explained as a smooth quiescent component disturbed by the
expansion of a bubble. That the length of the gas wisp exceeds by much the
thickness of the molecular disk of the Milky Way is consistent with the
cloud-formation scenario in which the gas is cold prior to the formation of
molecular clouds. Star formation in the filamentary gas wisp occurs at the edge
of a bubble (G52L nebula), which is consistent with some models of triggered
star formation.

Full text

arXiv:1310.3267v1 [astro-ph.GA] 11 Oct 2013
Astronomy & Astrophysics manuscript no. paper_filament_bubble c
ESO 2013
December 16, 2013
A 500 pc filamentary gas wisp in the disk of the Milky Way
Guang-Xing Li1, Friedrich Wyrowski1, Karl Menten1, and Arnaud Belloche1
Max-Planck Institut für Radioastronomie, Auf dem Hügel, 69, 53121 Bonn, Germany
December 16, 2013
ABSTRACT
Star formation occurs in molecular gas. In previous studies, the structure of the molecular gas has been studied in
terms of molecular clouds, but has been overlooked beyond the cloud scale. We present an observational study of the
molecular gas at 49.5◦< l < 52.5◦and −5.0 km s−1< vlsr <17.4 km s−1. The molecular gas is found in the form
of a huge (&500 pc)filamentary gas wisp. This has a large physical extent and a velocity dispersion of ∼5 km s−1.
The eastern part of the filamentary gas wisp is located ∼130 pc above the Galactic disk (which corresponds to 1.5–4
e-folding scale-heights), and the total mass of the gas wisp is &1×105M⊙. It is composed of two molecular clouds
and an expanding bubble. The velocity structure of the gas wisp can be explained as a smooth quiescent component
disturbed by the expansion of a bubble. That the length of the gas wisp exceeds by much the thickness of the molecular
disk of the Milky Way is consistent with the cloud-formation scenario in which the gas is cold prior to the formation of
molecular clouds. Star formation in the filamentary gas wisp occurs at the edge of a bubble (G52L nebula), which is
consistent with some models of triggered star formation.
Key words. ISM: clouds –ISM: bubbles–ISM: kinematics and dynamics —ISM: clouds– Stars: formation –Galaxies:
structure
1. Introduction
Molecular clouds belong to the densest and cold-
est parts of the Milky Way interstellar medium
(Field et al. 1969; McKee & Ostriker 1977). Shielded
from interstellar radiation fields, they provide con-
ditions necessary for star formation to take place.
Observationally, molecular clouds exhibit a compli-
cated, irregular, and filamentary morphology (Bally et al.
1987; Schneider & Elmegreen 1979; Williams et al. 2000;
Goldsmith et al. 2008; Men’shchikov et al. 2010), and
(sub)millimeter-line observations of molecular clouds sug-
gest that the gas in the clouds is moving supersonically.
Consensus has not been reached concerning the origin and
nature of molecular clouds.
It must be recognized that molecular gas is just one
of the phases of the Milky Way interstellar medium, and
its evolution is determined by many processes that occur in
the disk. To understand it, we must also look into the large-
scale structure of the multi-phased interstellar medium, and
understand the cloud evolution within this context.
Both observational and theoretical approaches have
been taken in this direction. Observationally, the distribu-
tion of molecular gas in nearby galaxies can be accessed
through millimeter line mapping (e.g. Schinnerer et al.
2013), from which structures such as spiral arms, fil-
aments, and spurs can be identified. Theoretically, the
structure of the multi-phased interstellar medium in a
galactic disk has been studied through simulating the
whole disk with different approaches (Kim & Ostriker 2002;
Shetty & Ostriker 2006; Tasker 2011; Dobbs et al. 2011;
Send offprint requests to: Guang-Xing Li, e-mail:
gxli@mpifr-bonn.mpg.de
Van Loo et al. 2013), complemented by analytical calcu-
lations (Lee & Shu 2012). It was found that filaments or
spurs can be created through the combination of gravita-
tional instability of a galactic disk, galactic shear, and fre-
quent encounters/agglomeration between molecular clouds
(Pringle et al. 2001; Dobbs & Pringle 2013).
In the Milky Way, studies of the structure of molecu-
lar gas have been confined spatially to the molecular cloud
scale or have been limited to the structure of the spiral
arms (Larson 1981; Solomon et al. 1987; Rathborne et al.
2009; Roman-Duval et al. 2009, 2010). This is partly due
to the complicated morphology of molecular gas and partly
due to the superposition of the emission of molecular gas
from different structures along the line of sight. In spite of
these difficulties, it is of both observational and theoreti-
cal interest to identify large, coherent molecular structures
in the Milky Way apart from the spiral arms, since these
structures are natural tracers of the large-scale gas circula-
tion in the Milky Way disk. In contrast to the extragalactic
case, where we are limited by the resolution and sensitiv-
ity of the telescopes (and the filtering of interferometers),
for our Milky Way it is possible to study the kinematics of
the molecular gas and the associated star formation with
in more detail.
The Milky Way interstellar medium has long been
thought to be dynamic. Shells and rims are generally
found in the disks of the Milky Way and other galaxies
(Churchwell et al. 2006). It has been proposed that the ex-
pansion of H i i regions, which creates shells and rims, can
collect the interstellar medium into a gravitationally unsta-
ble state (Elmegreen & Lada 1977; Whitworth et al. 1994;
Whitworth & Francis 2002), and trigger star formation.
The expansion of the bubbles can also energize the interstel-
Article number, page 1 of 9

A&A proofs: manuscript no. paper_filament_bubble
lar medium of the Milky Way efficiently (Norman & Ferrara
1996; Mac Low & Klessen 2004).
In this work, we present an observational study of the
region at 49.5◦< l < 52.5◦in the Milky Way. The molecu-
lar gas in the region exhibits a high degree of coherence, and
forms a filamentary gas wisp (gas filament) with a length
of ∼3◦. The eastern part of the filamentary gas wisp sits
at the edge of a bubble and is located at ∼0.75◦above
the galactic plane. This eastern part is listed in the context
of infrared bubbles as one of the “favorites of the Milky-
Way-Project volunteers” (Simpson et al. 2012), and it was
studied in terms of the G52L nebula by Bania et al. (2012),
who claimed that it may be the largest single H ii region in
the Milky Way. Based on several estimations (Watson et al.
2003; Anderson & Bania 2009; Roman-Duval et al. 2010;
Bania et al. 2012), the filamentary gas wisp has a distance
of 9.77 kpc, which implies a physical length of &500 pc.
This is ∼5times longer than the Nessie Nebula reported by
Jackson et al. (2010) 1. The physical length of the filamen-
tary gas wisp exceeds by much the size of a molecular cloud,
and this filamentary gas wisp is by far the largest coherent
molecular structure identified in the Milky Way. It exhibits
a coherent velocity structure, and is composed of several
molecular structures, including two molecular clouds and
one expanding bubble structure. We present observations
and an analysis of the region (Sect. 2, 3), followed by a
detailed discussion focusing on the implications on the dy-
namics of the Milky Way interstellar medium and the life
cycle of molecular gas (Sect. 4). In Sect. 5 we conclude.
2. Archival data
We obtained 3.6 µm and 8 µm data from the GLIMPSE
project (Benjamin et al. 2003), which is a fully sampled,
confusion-limited, four-band near-to-mid infrared survey of
the inner Galactic disk. We obtained 24 µm data from the
MIPSGAL project (Carey et al. 2009), which is a survey of
the Galactic disk with the MIPS instrument on Spitzer at
24 µm and 70 µm.
We obtained 13 CO(1-0) molecular line data (ν0=
110.2 GHz) from the Galactic Ring Survey (Jackson et al.
2006), which is a survey of the Milky Way disk with the
SEQUOIA multipixel array on the Five College Radio As-
tronomy Observatory 14 m telescope, and covers a longi-
tude range of 18◦< l < 55.7◦and a latitude range of
|b|<1◦with a spatial resolution of 46′′ .
3. Results
3.1. Region
Figure 1 shows the Spitzer three-color image of the re-
gion from 49.5◦< l < 52◦. The overlaid contours show
the molecular gas traced by 13 CO(1-0). The CO emis-
sion in all the panels is integrated within −5.0 km s−1<
vlsr <17.4 km s−1. Several features can be identified. At
51.5◦< l < 52.5◦, there is a bubble with a radius of ∼1◦
(G52L nebula, Bania et al. 2012). The molecular gas is sit-
uated to the north of the bubble. At 51◦< l < 52.5◦,
the molecular gas is organized in the form of two molecu-
lar clouds (G052.24+00.74 and G051.69+00.74). The cloud
1Note that Goodman et al. (2013) report a much larger length
of “many hundreds of pc” for the Nessie nebula. See also
http://milkywaybones.org/ for more details.
G052.24+00.74 has a roundish shape. This cloud is con-
nected with another molecular cloud, G051.69+00.74. This
cloud has a more elongated geometry, and star formation
occurs only at its eastern part.
At 49.5◦< l < 50◦, there is noticeable contamination
from gas with 5.2 km s−1< vlsr <7.2 km s−1(see the
red arrows in Fig. A.1). The contaminating gas has an ex-
tremely narrow line width (.0.5 km s−1) and tends to
spread along the spatial direction. This makes it easily dis-
tinguishable from the emission from the gas filament. This
narrow line width implies that the emission comes from
a close-by cloud. This is supported by the fact that the
contaminating gas has a more diffuse morphology (see Ap-
pendix A for 13 CO(1-0) channel maps of the region). This
distinction is similar to the supernova remnant G016.05-
0.57 studied in Beaumont et al. (2011).
The two molecular clouds (G052.24+00.74 and
G051.69+00.74) have a similar velocity and velocity disper-
sion: the cloud G052.24+00.74 has vlsr ∼4.6 km s−1and
δv ∼2.6 km s−1and the cloud G051.69+00.74 has vlsr ∼
3.6 km s−1and δv ∼3.6 km s−1(Roman-Duval et al.
2010). In the position-position and position-velocity space,
they are connected with some wispy gas filaments (at
l∼52◦and b∼0.8◦of the top and middle panels of Fig. 1).
The similarity of the velocity dispersions and the proximity
of the clouds in position-velocity space imply that the two
clouds are physically connected.
It can be readily seen from the 13 CO(1-0) emission that
this double-cloud system belongs to a large filament (Fig.
1 and Appendix A). The filamentary gas wisp is coher-
ent in both the spatial and the velocity direction, which
makes it distinguishable from other molecular structures,
for instance the ∼50 km s−1clouds (Appendix B). Seen
from the middle panel of Fig. 1, the filamentary gas wisp
extends from l= 49.5◦to l= 52.5◦, which implies an an-
gular extent of &3◦. Seen from the bottom panel of Fig.
1, the filamentary gas wisp has a limited velocity range of
∼22 km s−1(−5.0 km s−1< vlsr <17.4 km s−1). Sim-
ilar to the double-cloud system, all the molecular gas in
the filamentary gas wisp has a similar velocity dispersion.
At 49.5◦< l < 50.5◦, the filamentary gas wisp seems to
be split in both the position-position and position-velocity
maps (middle and bottom panels of Fig. 2). This coincides
with the presence of a bubble in the infrared band. To sum-
marize, the filamentary gas wisp is composed of two molec-
ular clouds and one bubble.
Star formation occurs in different parts of the filamen-
tary gas wisp. Star formation in molecular clouds can be
conveniently traced by 24 µm emission, which originates
from the dust heated by newly-born stars. In the Spitzer
image, this appears as red regions (Fig. 1 top and Fig. 3).
In the cloud pair G052.24+00.74 and G051.69+00.74, sev-
eral star-forming sites can be identified (Fig. 3) based on the
Spitzer 24 µm emission, three of which are currently host-
ing compact H i i regions (Lockman 1989; Urquhart et al.
2009). At 51.5◦< l < 52.5◦, all the star-forming sites are
located at the edge of the G52L bubble.
3.2. Distance and size of the filament
The distance to the region has been estimated by several
authors. Without trigonometric parallaxes, the distance to
the region can be determined with the kinematic method.
Article number, page 2 of 9

Guang-Xing Li et al.: 500 pc Filamentary Gas Wisp
Fig. 1. Top panel: Spitzer GLIMPSE (Benjamin et al. 2003) and MIPSGAL (Carey et al. 2009) three-color image of the region.
Red: 24 µm, green: 8 µm, blue: 3.6 µm. Overlaid contours are the velocity-integrated 13CO(1-0) emission (−5.0 km s−1< vlsr <
17.4 km s−1) from the Galactic Ring Survey (Jackson et al. 2006). Contours correspond to 3.5, 7.3, 11.2, 15 K km s−1.Middle
panel: Velocity-integrated 13CO(1-0) map of the region integrated within −4.95 km s−1< vlsr <17.36 km s−1. A scale bar of 500
pc is added assuming a kinematic distance of 9.8 kpc.Bottom panel: Galactic-latitude-integrated 13 CO(1-0) position-velocity map
of the region (integrated from −0.2◦< b < 1◦). The clouds G052.24+00.74, G051.69+00.74, the bubble at l∼50◦, and the G52L
nebula (Bania et al. 2012) are indicated in the middle and bottom panels. The velocity range we used to produce the 13 CO(1-0)
integrated intensity map is indicated in the bottom panel as the red shaded region. The emission at 5.2 km s−1< vlsr <7.2 km s−1
is due to contamination from a different molecular cloud, and some of the emission lies on the red dashed line. This component
has a smaller line width .0.5 km s−1, which implies that the contamination comes from a close-by cloud. This is supported by its
apparent diffuse morphology. Channel maps of the region are provided in Appendix A.
One key step in determining the kinematic distance is to
resolve the kinematic distance ambiguity.
There are different ways to resolve the ambiguity. Dis-
tance of the filamentary gas wisp can be determined by
studying the distance to molecular clouds and H ii re-
gions that belong to the filament. Using HI self-absorption,
Roman-Duval et al. (2010) found that the molecular clouds
G052.24+00.74 and G051.69+00.74 are located at the far
distance. Using the H2CO absorption line, Watson et al.
(2003) found that the H ii region G52.23+0.74 is located
at the far distance. Recently, Anderson & Bania (2009) and
Bania et al. (2012) studied the distance to the H i i re-
Article number, page 3 of 9

A&A proofs: manuscript no. paper_filament_bubble
Fig. 2. (a) Galactic longitude-latitude map of the peak temperatures of the 13 CO(1-0) data cube along the velocity axis. (b)
Galactic-longitude-velocity map of peak temperatures of 13 CO(1-0) along the galactic latitude axis. (c) Velocity-galactic-latitude
map of peak temperatures of 13 CO(1-0) along the galactic longitude axis. (d) Spitzer GLIMPSE (Benjamin et al. 2003) and
MIPSGAL (Carey et al. 2009) three-color image of the region. Red: 24 µm, Green: 8 µm, Blue: 3.6 µm. Overlaid contours are the
velocity-integrated 13CO(1-0) emission from the Galactic Ring Survey (Jackson et al. 2006) (−4.95 km s−1< vlsr <17.36 km s−1).
Contours correspond to 3.5, 7.3, 11.2, 15 K km s−1. In (a) (b), and (d), the bubble is indicated as a red ellipse.
gions G052.201+0.752 and G052.259+0.700 with HI emis-
sion/absorption method, and again found that they are at
the far distance. Therefore we conclude that the filamentary
gas wisp is located at the far distance, which is approxi-
mately 9.8 kpc. This suggests a galactocentric distance of
8.2 kpc, and the filamentary gas wisp probably resides in
or around the Perseus arm.
Accordingly, the filamentary gas wisp we identified has
a spatial extent of ∼500 pc. If the filamentary gas wisp
follows the spiral structure, it is probably angled ∼45◦
to our line of sight, and therefore probably has a depro-
Article number, page 4 of 9

Guang-Xing Li et al.: 500 pc Filamentary Gas Wisp
jected length a factor of √2longer. Therefore we conclude
that the filamentary gas wisp has a length of &500 pc. It
is one the of the largest coherent molecular structures in
the Milky Way apart from the spiral arms and the molec-
ular ring. The total mass of the filamentary gas wisp can
be estimated using the 13 CO(1-0) emission. To do this,
we integrated over the region with the line-of-sight inte-
grated flux I > 3.5 K km s−1. This corresponds to the first
contour in the upper panel of Fig. 1. This mass estimate
should be considered as a lower limit since by selecting this
threshold we only take the region with a high column den-
sity (NH2>1.75 ×1021 cm−2) into account. Using Eq. 1–3
of Roman-Duval et al. (2010) and assuming an excitation
temperature of 10 K, we obtain a total mass of ∼1×105M⊙
for the whole gas wisp (49.5◦< l < 52.5◦). According to
Simon et al. (2001), the derived mass is only weakly sen-
sitive to this choice of excitation temperature, and in our
case an excitation temperature of 20 K gives a mass of
∼0.6×105M⊙.
The two clouds at the eastern part of the filamentary
gas wisp have b∼0.74◦. Using a kinematic distance of
9.77 kpc, the double-cloud system is ∼130 pc above the
Galactic plane. At a Galactocentric distance of ∼8 kpc,
the molecular disk of the Milky Way has a FWHM thickness
of 90 −180 pc (at 7–8 kpc the FWHM is ∼90 pc and at 8–9
kpc the FWHM is 186 pc, Nakanishi & Sofue 2006). This
corresponds to an e-folding height of 38–80 pc. Therefore
the height of the double-cloud system is about 1.5–4 times
the e-folding height of the Galactic disk. The double-cloud
system is a unique cloud system that is located far above
the Galactic plane. According to Bania et al. (2012), one
possible explanation is that the material of the system has
been displaced by the expansion of the G52L nebula.
3.3. The bubble structure at l∼50◦
Figure 2 shows the bubble structure at l∼50◦. Its bound-
ary is visible in both the 8 µm emission, which traces poly-
cyclic aromatic hydrocarbon (PAHs) and in the 13 CO(1-0)
emission. The bubble is not easily visible at 24 µm, which
traces hot dust heated by a central star. Because of the ap-
parent absence of the 24 µm emission, the bubble structure
does not seem to be driven by the expansion of a H i i re-
gion. This is also supported by the absence of a diffuse H ii
region in the VGPS (Stil et al. 2006) continuum image.
It is more likely that the bubble structure is driven by
the expansion of a supernova. The Spitzer image of the
bubble resembles that of several supernova remnants in the
Churchwell et al. (2006) catalog. From panel (a) of Fig. 2,
using the kinematic distance of 9.8 kpc, we estimate a di-
ameter of 1◦∼180 pc, and from panel (b) of Fig. 2 we
estimate a total expansion velocity of 10 km s−1. These
give an age of 50 Myr. The energy of a possible super-
nova explosion can be estimated through the Sedov-Taylor
solution: E∼r3v2ρ∼0.16 ×1051 erg. The energy is
consistent with a supernova explosion. Here, a density of
10−24 g cm−3is used, which is typical of warm neutral
medium (cf. Bocchino et al. 2010).
4. Discussion
4.1. Morphology of the filamentary gas wisp
This giant molecular structure is among the largest molec-
ular structures studied in the Milky Way (&500 pc).
The physical size of the gas filamentary gas wisp is much
larger than that of a typical molecular cloud (∼10 pc,
Roman-Duval et al. 2010). The velocity dispersion along a
single line of sight in the filamentary gas wisp is not signif-
icantly different from that of ordinary molecular clouds.
The molecular gas in the filamentary gas wisp is concen-
trated in the vertical (Galactic-latitude) direction and elon-
gated along the horizontal (Galactic-longitude) direction.
At different locations, the filamentary gas wisp exhibits a
different width. At 49.5◦< l < 50.5◦, the filamentary gas
wisp is split in the map, which makes it difficult to define
its width. From the map, the cloud G052.24+00.74 appears
to be more extended in the vertical direction than the cloud
G051.69+00.74. We therefore used its vertical extent as an
estimate of the width of the filamentary gas wisp. The ver-
tical extent of the cloud G051.69+00.74 is measured for the
region with I > 3.5 K km s−1(NH2>1.75 ×1021 cm−2).
This corresponds to the first contour in the upper panel of
Fig. 1. We found that the cloud extends from b∼0.65◦
to b∼0.82◦. From this we estimated a diameter of 30 pc,
which implies an aspect ratio of ∼600/30 = 20 for the gas
wisp. The gas wisp is one of the most elongated molecu-
lar structures found in the Milky Way (see also the Nessie
nebula, Jackson et al. 2010). The width of the filamen-
tary gas wisp is narrower than the FWHM thickness of the
molecular disk of the Milky Way, which is about 90–180
pc (Nakanishi & Sofue 2006) at a Galactocentric distance
of ∼8kpc.
Similar large-scale molecular structures have been ob-
served in other galaxies. In spiral galaxies, elongated gas
condensations are frequently observed. They can be seen as
narrow dark lanes that extend perpendicular to the spiral
arms (Lynds 1970; Weaver 1970). The exact definitions of
dust lanes or spurs differ in the literature. However, in most
cases spurs refer to the objects whose widths are similar to
that of spiral arms (Elmegreen 1980). In our case, the fila-
mentary gas wisp should not be termed a spur because its
width is narrower than the width of the spiral arms of a
typical galaxy, which is ∼500 pc (Egusa et al. 2011).
In our case, the filamentary gas wisp is about one or
two orders of magnitudes longer than ordinary molecular
clouds, but is still much narrower than the spurs in galaxies.
Therefore we propose that the filamentary gas wisp is a
new object that is yet to be classified. Because of this, we
termed it a gas wisp in this work to emphasize its elongated
morphology.
Even though the thickness of the filamentary gas
wisp is similar to the resolution of the PAWS survey
(Schinnerer et al. 2013) of M51, filamentary gas wisps of
this size would not be detected. This is because the survey
is only sensitive to objects with a mass &1.2×105M⊙and
the clouds in the filamentary gas wisp are only ∼104M⊙.
However, similar large-scale gas structures in nearby galax-
ies are probably suitable targets for ALMA thanks to its im-
proved sensitivity. Nearby face-on galaxies are expected to
be excellent sites for studying these gas wisps since line-of-
sight confusion can be avoided. A future pro ject at ALMA
targeting at the molecular gas in nearby face-on galaxies is
expected to resolve similar gas condensations and provide
Article number, page 5 of 9

A&A proofs: manuscript no. paper_filament_bubble
a more complete picture of the structure of molecular gas
in galaxies.
4.2. Implications for the formation of molecular clouds
The formation and evolution of molecular clouds is one of
the fundamental problems in interstellar medium studies.
To account for the short formation timescale of molecular
clouds, two scenarios have been proposed. The first sce-
nario involves colliding flows. In this scenario, molecular
clouds form from diffuse gas (warm neutral medium) col-
lected into a dense phase (cold neutral medium/molecular
medium) through colliding flows (Audit & Hennebelle
2005; Heitsch et al. 2006; Vázquez-Semadeni et al. 2007,
2010; Inoue & Inutsuka 2012). The molecular gas can form
quickly in the converging flows because of dynamically-
triggered thermal instability.
The second scenario has been proposed by Pringle et al.
(2001) and Dobbs & Pringle (2013). In this scenario, the
gas is already relatively dense and cold prior to becoming
a giant molecular cloud. According to Pringle et al. (2001),
there is expected to be copious cold gas in the inter-arm
regions since the circulation of the molecular gas is a
process that occurs at the disk scale. To describe the global
circulation of the gas, we divided it into the in-arm phase
in which the gas is situated inside the spiral arms, and the
inter-arm phase in which the gas is situated in the inter-
arm regions. As discussed in Pringle et al. (2001), in the
inter-arm phase, the cold gas exists in the form of wisps,
and the gas in these wisps will show up as giant molecular
clouds during the spiral-arm phase. This has been largely
confirmed by the simulations of Dobbs & Pringle (2013),
which track the evolution of single molecular clouds. These
authors found that molecular clouds begin to disperse as
they leave the spiral arm. Due to differential shear, the
molecular clouds are transformed into filamentary gas
wisps in the inter-arm region. Since the shear occurs at a
large scale, we expect to see gas wisps whose physical scale
exceeds the thickness of the Milky Way disk. In our case,
the physical length of the filamentary gas wisp (&500
pc) is much larger than the scale-height of the Milky Way
molecular disk. This is consistent with the cloud-formation
scenario by Pringle et al. (2001) and Dobbs & Pringle
(2013). Such large-scale structures are also observed
in other numerical simulations of galactic disks (Tan
2000; Kim & Ostriker 2002; Dobbs & Bonnell 2006;
Dobbs et al. 2006; Shetty & Ostriker 2006; Tasker & Tan
2009; Ceverino et al. 2012).
On the other hand, it is difficult to understand the
filamentary gas wisp in the converging flow scenario. In
this scenario, molecular gas forms from the converging
HI gas through the dynamically-triggered thermal insta-
bility. As summarized in Dobbs et al. (2012), the sources
of the converging flows can be stellar winds or su-
pernovae (Koyama & Inutsuka 2000; Heitsch & Hartmann
2008; Ntormousi & Burkert 2011), turbulence in the in-
terstellar medium (Ballesteros-Paredes et al. 1999), spiral
shocks (Leisawitz & Bash 1982), and gravitational instabil-
ity. In our case, the filamentary gas wisp cannot be created
by converging stellar winds, supernovae, or turbulence in
the interstellar medium, since these mechanisms are local
and cannot create structures that are larger than the thick-
ness of the Milky Way disk. Spiral shocks and gravitational
instability might create conditions favorable for converging
flows to occur. However, to access these possibilities we need
to simulate converging flows in a galactic context and prop-
erly quantify the role of the dynamically-triggered thermal
instability in the formation of molecular gas. This task has
not been achieved yet. Either the converging flow scenario
is unable to explain how filamentary gas wisps form, or
our current understanding of converging flows in a galactic
context is incomplete.
4.3. Star formation in the molecular cloud pair
G0524.2+00.74 and G051.69+00.74
It is unclear to what extent molecular clouds are grav-
itationally bound. Gravity is important at a variety
of physical scales during star formation. According to
Roman-Duval et al. (2010), the clouds G052.24+00.74 and
G051.69+00.74 have virial parameters of 0.29 and 0.33, re-
spectively. This means that both clouds are gravitationally
bound 2.
Star formation takes place in several sites, which is
traced by the 24 µm emission in the Spitzer image. As in-
dicated in Fig. 3, all these sites seem to be located at the
edge of a bubble (G52L nebula).
The connection between the location of the star-forming
sites and the edge of the bubble agrees with the sta-
tistical study of Thompson et al. (2012), in which a sig-
nificant overdensity of young stellar objects toward the
edges of the bubbles was found. This is consistent with
the collect-and-collapse scenario of triggered star formation
(Elmegreen & Lada 1977; Whitworth et al. 1994).
5. Conclusions
We studied a giant coherent molecular structure (a filamen-
tary gas wisp) at 49.5◦< l < 52.5◦. The eastern part of the
filamentary gas wisp is located ∼130 pc above the Galac-
tic disk (which corresponds to 1.5–4 e-folding scale-heights),
and the total mass of the gas wisp is &1×105M⊙. Apart
from the spiral arms and the molecular ring, this is among
the largest coherent molecular structures identified in the
Milky Way. The velocity structure of the filamentary gas
wisp is coherent and smooth at 50.5◦< l < 52.5◦, and at
49.5◦< l < 50.5◦, the gas wisp is disturbed by a bubble
structure. This might be caused by a supernova. The over-
all velocity structure of the filamentary gas wisp can be
understood as a quiescent filamentary gas wisp disturbed
by the expansion of a bubble. The eastern part of the fila-
mentary gas wisp is composed of a system of two molecular
clouds (G052.24+00.74 and G051.69+00.74) and is located
∼130 pc above the Galactic plane.
Star formation already takes place in several parts of
this filamentary gas wisp. In the cloud pair G052.24+00.74
and G051.69+00.74, nearly the entire star formation oc-
curs at the edge of a bubble (G52L nebula, Bania et al.
2012). This is consistent with the collect-and-collapse sce-
nario of triggered star formation (Elmegreen & Lada 1977;
Whitworth et al. 1994), and can be understood in the sta-
tistical context of Thompson et al. (2012).
The discovery of this filamentary gas wisp, whose
length exceeds the thickness of the molecular disk of the
2Here and in Roman-Duval et al. (2010) the virial parameter
αof a molecular cloud is the ratio of its virial mass Mvir to its
mass.
Article number, page 6 of 9

Guang-Xing Li et al.: 500 pc Filamentary Gas Wisp
51 .2
51 .4
51 .651 .852 .052 .252 .4
Gal Lon (De gre e)
0. 5
0. 6
0. 7
0. 8
0. 9
Gal Lat (D egr ee)
G052.24+00.74 G051.69+00.74
G52 Nebula
Fig. 3. Spitzer GLIMPSE (Benjamin et al. 2003) and MIPSGAL three-color image of the clouds G052.24+00.74 and G51.69+00.74.
Red: 24 µm, green: 8 µm, blue: 3.6 µm. Overlaid contours are the velocity-integrated 13CO(1-0) emission (−4.95 km s−1< vlsr <
17.36 km s−1) from the Galactic Ring Survey (Jackson et al. 2006). Contours correspond to 3.5, 7.3, 11.2, 15 K km s−1. The cloud
G052.24+00.74 and G51.69+00.74 as well as the G52L nebula are indicated.
Milky Way, suggests that the formation and evolution
of molecular clouds is a phenomenon that occurs at the
disk scale. The large physical extent is consistent with
the cloud-formation scenario by Pringle et al. (2001) and
Dobbs & Pringle (2013), in which the gas that constitutes
the molecular clouds is already relatively cold prior to the
cloud formation.
We are currently unable to answer how representative
this filamentary gas wisp is in the Milky Way disk. One
reason is that we are restricted by the line-of-sight confu-
sion, and the filamentary gas wisp is fragile in nature. In
our case, at 49.5◦< l < 50.5◦, the filamentary gas wisp
is already being destroyed by the expansion of a bubble
structure. It is possible that a significant fraction of the
gas in the Milky Way exists in this form during at least
part of its lifetime. Another difficulty is to properly quan-
tify the coherence of molecular structures beyond the cloud
scale. A position-velocity plot of the 13 CO(1-0) data from
the same region shown in the bottom panel of Fig. 1 re-
veals filamentary structures at vlsr ∼50 km s−1. From
a visual inspection we found that these structures are not
as coherent as the filamentary gas wisp (see Appendix B
for a comparison). In general one cannot yet quantify the
coherence of molecular structures in the Milky Way. More
studies of the morphology of the molecular gas in both the
Milky Way and other galaxies with improved observations
and analyses are needed to fully understand the circulation
of molecular gas at large scales.
Acknowledgements. Guang-Xing Li is supported for this research
through a stipend from the International Max Planck Research School
(IMPRS) for Astronomy and Astrophysics at the Universities of Bonn
and Cologne. This publication makes use of molecular line data from
the Boston University-FCRAO Galactic Ring Survey (GRS). The
GRS is a joint project of Boston University and Five College Ra-
dio Astronomy Observatory, funded by the National Science Founda-
tion under grants AST-9800334, AST-0098562, & AST-0100793. This
work is based in part on observations made with the Spitzer Space
Telescope, which is operated by the Jet Propulsion Laboratory, Cali-
fornia Institute of Technology under a contract with NASA. We thank
James Urquhart and Malcolm Walmsley for careful readings of our pa-
per and for many insightful comments. We thank the referee Adam
Ginsburg for several thorough and careful reviews of the paper and
for his insightful comments.
Appendix A: Channel map of the filamentary gas
wisp
In Fig. A.1 we present the channel maps of the 13 CO(1-0)
emission from the GRS (Jackson et al. 2006) survey. Some
contamination from local clouds is indicated.
Appendix B: A comparison with CO emission from
the ∼50 km s−1component
To demonstrate the coherence of our filamentary gas wisp,
in Fig. B.1 we present a map of the same region with the ve-
locity integrated within 29.5 km s−1< vlsr <73.3 km s−1.
Seen from the 13 CO(1-0) emission, this component is com-
posed of individual patches of molecular clouds and is not
as coherent as our filamentary gas wisp. The distance to
the region is ∼5.3kpc.
References
Anderson, L. D. & Bania, T. M. 2009, ApJ, 690, 706
Audit, E. & Hennebelle, P. 2005, A&A, 433, 1
Ballesteros-Paredes, J., Hartmann, L., & Vázquez-Semadeni, E. 1999,
ApJ, 527, 285
Bally, J., Lanber, W. D., Stark, A. A., & Wilson, R. W. 1987, ApJ,
312, L45
Bania, T. M., Anderson, L. D., & Balser, D. S. 2012, ApJ, 759, 96
Beaumont, C. N., Williams, J. P., & Goodman, A. A. 2011, ApJ, 741,
14
Benjamin, R. A., Churchwell, E., Babler, B. L., et al. 2003, PASP,
115, 953
Bocchino, F., Bandiera, R., & Gelfand, J. 2010, A&A, 520, A71
Carey, S. J., Noriega-Crespo, A., Mizuno, D. R., et al. 2009, PASP,
121, 76
Ceverino, D., Dekel, A., Mandelker, N., et al. 2012, MNRAS, 420,
3490
Churchwell, E., Povich, M. S., Allen, D., et al. 2006, ApJ, 649, 759
Dobbs, C. L. & Bonnell, I. A. 2006, MNRAS, 367, 873
Dobbs, C. L., Bonnell, I. A., & Pringle, J. E. 2006, MNRAS, 371, 1663
Article number, page 7 of 9

A&A proofs: manuscript no. paper_filament_bubble
0.25
0.50
0.75
1.00
b
0.25
0.50
0.75
1.00
b
0.25
0.50
0.75
1.00
b
0.25
0.50
0.75
1.00
b
0.25
0.50
0.75
1.00
b
0.25
0.50
0.75
1.00
b
4849505152
l
4849505152
l
0
1
T [ K]
Fig. A.1. Channel maps of the GRS (Jackson et al. 2006) 13 CO(1-0) emission of the region. The contaminating gas has a narrow
(∼0.5 km s−1) line width, and is detected only in single channels in the map. Some contamination from local clouds is indicated
with red circles as examples, and these red circles are also plotted in yellow in the adjacent velocity channels for comparison.
Fig. B.1. Velocity-integrated 13CO(1-0) map of the ∼50 km s−1clouds. The emission is from the same region as shown in Fig. 1
and is integrated within 29.5 km s−1< vlsr <73.3 km s−1.
Dobbs, C. L., Burkert, A., & Pringle, J. E. 2011, MNRAS, 413, 2935
Dobbs, C. L. & Pringle, J. E. 2013, MNRAS, 432, 653
Dobbs, C. L., Pringle, J. E., & Burkert, A. 2012, MNRAS, 425, 2157
Egusa, F., Koda, J., & Scoville, N. 2011, ApJ, 726, 85
Elmegreen, B. G. & Lada, C. J. 1977, ApJ, 214, 725
Elmegreen, D. M. 1980, ApJ, 242, 528
Field, G. B., Goldsmith, D. W., & Habing, H. J. 1969, ApJ, 155, L149
Goldsmith, P. F., Heyer, M., Narayanan, G., et al. 2008, ApJ, 680,
428
Goodman, A. A., Alves, J. F., Beaumont, C., et al. 2013, in Amer-
ican Astronomical Society Meeting Abstracts, Vol. 221, American
Astronomical Society Meeting Abstracts, 234.01
Heitsch, F. & Hartmann, L. 2008, ApJ, 689, 290
Heitsch, F., Slyz, A. D., Devriendt, J. E. G., Hartmann, L. W., &
Burkert, A. 2006, ApJ, 648, 1052
Inoue, T. & Inutsuka, S.-i. 2012, ApJ, 759, 35
Jackson, J. M., Finn, S. C., Chambers, E. T., Rathborne, J. M., &
Simon, R. 2010, ApJ, 719, L185
Jackson, J. M., Rathborne, J. M., Shah, R. Y., et al. 2006, ApJS, 163,
145
Kim, W.-T. & Ostriker, E. C. 2002, ApJ, 570, 132
Koyama, H. & Inutsuka, S.-I. 2000, ApJ, 532, 980
Larson, R. B. 1981, MNRAS, 194, 809
Lee, W.-K. & Shu, F. H. 2012, ApJ, 756, 45
Leisawitz, D. & Bash, F. 1982, ApJ, 259, 133
Lockman, F. J. 1989, ApJS, 71, 469
Lynds, B. T. 1970, in IAU Symposium, Vol. 38, The Spiral Structure
of our Galaxy, ed. W. Becker & G. I. Kontopoulos, 26
Mac Low, M.-M. & Klessen, R. S. 2004, Reviews of Modern Physics,
76, 125
Article number, page 8 of 9

Guang-Xing Li et al.: 500 pc Filamentary Gas Wisp
McKee, C. F. & Ostriker, J. P. 1977, ApJ, 218, 148
Men’shchikov, A., André, P., Didelon, P., et al. 2010, A&A, 518, L103
Nakanishi, H. & Sofue, Y. 2006, PASJ, 58, 847
Norman, C. A. & Ferrara, A. 1996, ApJ, 467, 280
Ntormousi, E. & Burkert, A. 2011, ArXiv e-prints
Pringle, J. E., Allen, R. J., & Lubow, S. H. 2001, MNRAS, 327, 663
Rathborne, J. M., Johnson, A. M., Jackson, J. M., Shah, R. Y., &
Simon, R. 2009, ApJS, 182, 131
Roman-Duval, J., Jackson, J. M., Heyer, M., et al. 2009, ApJ, 699,
1153
Roman-Duval, J., Jackson, J. M., Heyer, M., Rathborne, J., & Simon,
R. 2010, ApJ, 723, 492
Schinnerer, E., Meidt, S. E., Pety, J., et al. 2013, ArXiv e-prints
Schneider, S. & Elmegreen, B. G. 1979, ApJS, 41, 87
Shetty, R. & Ostriker, E. C. 2006, ApJ, 647, 997
Simon, R., Jackson, J. M., Clemens, D. P., Bania, T. M., & Heyer,
M. H. 2001, ApJ, 551, 747
Simpson, R. J., Povich, M. S., Kendrew, S., et al. 2012, MNRAS, 424,
2442
Solomon, P. M., Rivolo, A. R., Barrett, J., & Yahil, A. 1987, ApJ,
319, 730
Stil, J. M., Taylor, A. R., Dickey, J. M., et al. 2006, AJ, 132, 1158
Tan, J. C. 2000, ApJ, 536, 173
Tasker, E. J. 2011, ApJ, 730, 11
Tasker, E. J. & Tan, J. C. 2009, ApJ, 700, 358
Thompson, M. A., Urquhart, J. S., Moore, T. J. T., & Morgan, L. K.
2012, MNRAS, 421, 408
Urquhart, J. S., Hoare, M. G., Purcell, C. R., et al. 2009, A&A, 501,
539
Van Loo, S., Butler, M. J., & Tan, J. C. 2013, ApJ, 764, 36
Vázquez-Semadeni, E., Colín, P., Gómez, G. C., Ballesteros-Paredes,
J., & Watson, A. W. 2010, ApJ, 715, 1302
Vázquez-Semadeni, E., Gómez, G. C., Jappsen, A. K., et al. 2007,
ApJ, 657, 870
Watson, C., Araya, E., Sewilo, M., et al. 2003, ApJ, 587, 714
Weaver, H. F. 1970, in IAU Symposium, Vol. 39, Interstellar Gas
Dynamics, ed. H. J. Habing, 22
Whitworth, A. P., Bhattal, A. S., Chapman, S. J., Disney, M. J., &
Turner, J. A. 1994, A&A, 290, 421
Whitworth, A. P. & Francis, N. 2002, MNRAS, 329, 641
Williams, J. P., Blitz, L., & McKee, C. F. 2000, Protostars and Planets
IV, 97
Article number, page 9 of 9