$&6.59&+1.$34+,2
"e Wise Catalog Of Galactic H Ii Regions
L. D. Anderson
T. M. Bania
Dana S. Balser
V. Cunningham
T. V. Wenger
See next page for additional authors
1..185+,4$0'$'',5,10$.813-4$5 +<243(4($3&+3(214,5139876('6)$&6.59#26%.,&$5,104
;,435,&.(,4%316*+551916)13)3(($0'12(0$&&(44%9;((4($3&+(214,5139"! 5+$4%((0$&&(25(')13,0&.64,10,0$&6.59&+1.$34+,2
%9$0$65+13,:('$'/,0,453$5131);((4($3&+(214,5139"! 13/13(,0)13/$5,102.($4(&105$&5 ,$0+$3/10/$,.876('6
,*,5$.1//104,5$5,10
0'(3410$0,$$.4(3$0$600,0*+$/!"(0*(3!1+04510($0'3/(053165";(",4(
$5$.1*)$.$&5,&,(*,104 Faculty Scholarship
+<243(4($3&+3(214,5139876('6)$&6.59#26%.,&$5,104
The Astrophysical Journal Supplement Series, 212:1 (18pp), 2014 May doi:10.1088/0067-0049/212/1/1
C
2014. The American Astronomical Society. All rights reserved. Printed in the U.S.A.
THE WISE CATALOG OF GALACTIC H ii REGIONS
L. D. Anderson
1,5
,T.M.Bania
2
, Dana S. Balser
3
, V. Cunningham
1
, T. V. Wenger
3,4
,
B. M. Johnstone
1
, and W. P. Armentrout
1
1
Department of Physics and Astronomy, West Virginia University, Morgantown, WV 26506, USA
2
Institute for Astrophysical Research, Department of Astronomy, Boston University,
725 Commonwealth Avenue, Boston, MA 02215, USA
3
National Radio Astronomy Observatory, 520 Edgemont Road, Charlottesville, VA 22903, USA
4
Department of Astronomy, University of Virginia, P.O. Box 3813, Charlottesville, VA 22904, USA
Received 2013 November 8; accepted 2013 December 20; published 2014 April 10
ABSTRACT
Using data from the all-sky Wide-Field Infrared Survey Explorer (WISE) satellite, we made a catalog of over
8000 Galactic H ii regions and H ii region candidates by searching for their characteristic mid-infrared (MIR)
morphology. WISE has sufficient sensitivity to detect the MIR emission from H ii regions located anywhere in the
Galactic disk. We believe this is the most complete catalog yet of regions forming massive stars in the Milky Way.
Of the ∼8000 cataloged sources, ∼1500 have measured radio recombination line (RRL) or Hα emission, and are
thus known to be H ii regions. This sample improves on previous efforts by resolving H ii region complexes into
multiple sources and by removing duplicate entries. There are ∼2500 candidate H ii regions in the catalog that are
spatially coincident with radio continuum emission. Our group’s previous RRL studies show that ∼95% of such
targets are H ii regions. We find that ∼500 of these candidates are also positionally associated with known H ii
region complexes, so the probability of their being bona fide H ii regions is even higher. At the sensitivity limits
of existing surveys, ∼4000 catalog sources show no radio continuum emission. Using data from the literature, we
find distances for ∼1500 catalog sources, and molecular velocities for ∼1500 H ii region candidates.
Key words: Galaxy: structure – H ii regions – infrared: ISM – ISM: bubbles – stars: formation
Online-only material: color figures, machine-readable tables
1. INTRODUCTION
H ii regions are the zones of ionized gas surrounding young
massive stars. The stars capable of producing the ultra-violet
photons necessary to ionize their surrounding medium have
spectral types of ∼B0 or earlier. Such stars only live ∼10 Myr
and thus H ii regions are zero-age objects compared to the age
of the Milky Way: they trace star formation at the present
epoch. H ii regions are the brightest objects in the Galaxy at
infrared (IR) and radio wavelengths and can be detected across
the entire Galactic disk. Unlike other tracers of Galactic star
formation, the identification of an H ii region unambiguously
locates massive star formation. They are the archetypical tracers
of spiral arms and have been instrumental in creating a better
understanding of the structure of our Galaxy. Their chemical
abundances represent Galactic abundances today, and reveal the
effects of billions of years of Galactic chemical evolution. They
are the main contributors to the ionized photons in a galaxy,
the emission from which is used to determine extragalactic and
Galactic star formation rates. In short, Galactic H ii regions
are extremely important objects for learning about a number
of problems in astrophysics, including star formation, Galactic
structure, and Galactic evolution.
Despite their importance, the census of Galactic H ii regions
is severely incomplete, as evidenced by the recent Green Bank
Telescope H ii
Region Discovery Survey (GBT HRDS; Bania
et al. 2010; Anderson et al. 2011). The GBT HRDS measured the
radio r ecombination line (RRL) and radio continuum emission
from 448 previously unknown Galactic H ii regions. Over the
survey zone, the GBT HRDS doubled the census of known H ii
5
Also Adjunct Astronomer at the National Radio Astronomy Observatory,
P.O. Box 2, Green Bank, WV 24944, USA
regions with measured RRL emission. The average on-source
integration time in the GBT HRDS was only ∼10 minutes;
Anderson et al. (2011) found hundreds more candidate H ii
regions that would have required longer integrations. This hints
at a larger population of Galactic H ii regions about which
nothing is known.
As the GBT HRDS demonstrated, Galactic H ii regions can be
easily and reliably identified from mid-infrared (MIR) data. If
the resolution of the MIR data is sufficient, all H ii regions have
essentially the same MIR morphology: their ∼10 μm emission
surrounds their ∼20 μm emission and the latter is coincident
with the ionized gas traced by radio continuum emission (see
Anderson et al. 2011). This characteristic morphology allows
one to identify H ii region candidates in MIR images. Radio
continuum and RRL observations can then confirm that these
targets are H ii regions. The identification of young H ii regions
in IR data, where the emission is from heated dust, has motivated
much of the H ii region research over the past 25 yr (e.g., Wood
& Churchwell 1989; Kurtz et al. 1994).
In the GBT HRDS, Anderson et al. (2011) used data from
the Spitzer legacy 24 μm MIPSGAL survey (Carey et al. 2009)
to identify targets. The Spitzer legacy surveys were generally
limited to within 1
◦
of the Galactic mid-plane, and || 65
◦
.
Most regions of massive star formation are within this Galactic
zone, but a complete sample of Galactic regions forming massive
stars requires coverage outside the zone surveyed by Spitzer.
Data from the all-sky Wide-Field Infrared Survey Explorer
(WISE) can also be used to identify H ii regions. WISE covers the
entire sky in four photometric bands: 3.4 μm, 4.6 μm, 12 μm,
and 22 μm at angular resolutions of 6.
1, 6.
4, 6.
5, and 12
,
respectively. H ii regions appear visually similar in the WISE
12 μm and 22 μm bands compared with the Spitzer IRAC (Fazio
et al. 2004)8.0μm and MIPS (Rieke et al. 2004)24μm bands,
1
The Astrophysical Journal Supplement Series, 212:1 (18pp), 2014 May Anderson et al.
1 10 100 1000 10000
MIPSGAL 24 μm flux (Jy)
0.1
1.0
10.0
100.0
VGPS 21 cm flux (Jy)
F
24
= 0.03 F
21cm
0.90
Figure 1. Correlation between 24 μm and 21 cm fluxes for Galactic H ii regions.
The data points are from a sample of 301 H ii regions from 15
◦
55
◦
,
|b| 1
◦
(Anderson 2010). The radio and MIR fluxes are highly correlated. The
larger scatter at lower fluxes is likely due to photometric errors and uncertainties
in the background estimation.
respectively (see Anderson et al. 2012b). For H ii regions, the
8.0 μm and 12 μm emission are both largely due to polycyclic
aromatic hydrocarbon (PAH) molecules, which fluoresce in
ultra-violet radiation fields. The IRAC 8.0 μm band contains
strong PAH emission at 7.7 μm and 8.6 μm, whereas the WISE
12 μm band contains PAH emission at 11.2 μm and 12.7 μm
(see Tielens 2008, for a review). The MIPS 24 μm band actually
has a very similar bandpass compared with the WISE 22 μm
band. For H ii regions this band traces stochastically heated
small dust grains (Watson et al. 2008, 2009; Deharveng et al.
2010; Anderson et al. 2012c). The WISE 22 μm band resolution
and sensitivity (12
and 6 mJy) are comparable to those of
MIPSGAL (6
and 1.3 mJy).
2. CREATING THE WISE CATALOG
OF GALACTIC H ii REGIONS
WISE can in principle detect the MIR emission from all
Galactic H ii regions. Anderson (2010) measured the integrated
MIPSGAL flux and 21 cm Very Large Array (VLA) Galactic
Plane Survey (VGPS; Stil et al. 2006) continuum emission from
a large sample of 301 first quadrant Galactic H ii regions. Using
these data, Figure 1 shows that the emission at 21 cm wavelength
is ∼30 times less than that at 24 μm. Anderson et al. (2012b)
found that the WISE 22 μm flux of H ii regions is the same as
the MIPSGAL 24 μm flux. The sensitivity of WISE,6mJyat
22 μm, is therefore able to detect H ii regions with integrated
21 cm fluxes of ∼0.2 mJy. We show in Figure 2 the expected
flux for H ii regions ionized by single stars of various spectral
types. For the calculation of the ionizing flux we used Sternberg
et al. (2003) and for the conversion from ionizing flux to 21 cm
luminosity we used the relation given in Rubin (1968). The
expected 21 cm flux for an H ii region at the sensitivity limit of
WISE, using the 30:1 ratio found for the ratio of the MIPSGAL
to VGPS fluxes, is well below that required to detect the MIR
emission from all Galactic H ii regions.
This holds true even when extinction is factored into the
calculations. Flaherty et al. (2007) find that for Spitzer the
extinction at 24 μm is about half that at 2.16 μm(K
S
band).
Because the 22 μm WISE filter is similar to the 24 μm Spitzer
filter, the 22 μm WISE band will share essentially the same
value. The extinction in the 2.16 μm K
S
band is nominally
about a tenth that of visual (Rieke & Lebofsky 1985). Therefore,
0 5 10 15 20 25 30
Distance (kpc)
10
-1
10
0
10
1
10
2
10
3
10
4
10
5
21 cm Flux density (mJy)
O3
O4
O5
O6
O7
O8
O9
B0
WISE
Figure 2. Model 21 cm flux densities for H ii regions ionized by single stars of
spectral types O3–B0 as a function o f distance from the Sun. For the calculation
of the ionizing flux we used Sternberg et al. (2003) and for the conversion from
ionizing flux to 21 cm luminosity we used the relation given in Rubin (1968).
The width of the curves reflects a range of nebular electron temperatures from
5000 K to 10,000 K. The vertical dotted line marks the most distant H ii region
currently known, ∼20 kpc. The horizontal dotted line is the expected radio flux
for an H ii region at the sensitivity limit of the 22 μm WISE data, ∼0.2mJy.
WISE has the sensitivity to detect all Galactic H ii regions.
(A color version of this figure is available in the online journal.)
A
24
/A
V
0.05 and even 50 mag of visual extinction would
only result in 2.5 mag of 22 μm extinction. Two and a half
magnitudes of extinction corresponds to an intensity decrease
of a factor of 10. Even after applying this factor, the WISE
sensitivity in Figure 2 is well below the flux from an H ii region
ionized by a single B0 star anywhere in the Galaxy. Furthermore,
due to the warp of t he Galactic disk, the most distant H ii regions
are generally found above and below the Galactic mid-plane,
where the line-of-sight extinction is lower. In sum, the MIR
line-of-sight extinction is sufficiently low so that we can detect
extremely distant H ii regions with WISE.
We use visual and automatic searches of WISE data to identify
H ii regions from their MIR emission morphology. We visually
search WISE 12 μm and 22 μm i mages spanning the entire
Galactic plane within 8
◦
of the nominal mid-plane, |b| 8
◦
.
We create WISE 12 μm and 22 μm mosaics using the Montage
software.
6
The WISE public image tiles are 1.
◦
564 × 1.
◦
564.
We combine these tiles into mosaics 4
◦
in longitude, and 16
◦
in latitude, centered on the Galactic plane. Adjacent mosaics
overlap in longitude by 0.
◦
5. Our WISE mosaics collectively
cover |b| 8
◦
over the entire range of Galactic longitudes. For
both photometric bands, we use a pixel size of 2
.TheWISE
tiles are individually background corrected so the image tile
6
http://montage.ipac.caltech.edu/
2
The Astrophysical Journal Supplement Series, 212:1 (18pp), 2014 May Anderson et al.
Table 1
Radio Continuum Surveys
Survey Wavelength Longitudes Used Latitudes Used Resolution Reference
a
(cm) (arcsec)
MAGPIS 6 350
◦
<<42
◦
|b| 0.
◦
441
MAGPIS 20 5
◦
<<48
◦
|b| 0.
◦
862
VGPS 21 18
◦
<<66
◦
|b| from 1
◦
to 2
◦
60 3
CGPS 21 66
◦
<<175
◦
−3.
◦
5 <b<+5.
◦
560 4
NVSS 20 66
◦
<<247
◦
|b| 8
◦
45 5
SGPS 21 253
◦
<<358
◦
|b| 1.
◦
5 120 6
SUMSS 36 247
◦
<<337
◦
|b| 8
◦
45 7
Note.
a
(1) Becker et al. 1994; (2) Helfand et al. 2006; (3) Stil et al. 2006; (4) Taylor et al. 2003; (5) Condon et al. 1998;
(6) McClure-Griffiths et al. 2005; (7) Bock et al. 1999.
fluxes are not on an absolute scale. We find that, when making
background corrections, the Montage software often introduces
large angular-scale (∼10
◦
) variations in the background level.
The quality of the mosaics is nevertheless sufficient for the
identification of sites of massive star formation.
To our knowledge, there are only five massive star formation
complexes known outside the latitude range of our WISE
mosaics (|b| 8
◦
): Mon R2, the California Nebula, Orion,
BFS 11, and G159.6−18.5 in the Perseus molecular cloud. For
these regions we create and search WISE mosaics as before and
all five are detected by WISE. The California Nebula (Sharpless
220) is a long filamentary structure s een at 12 μm. We include it
in the catalog, although its appearance is different from that of
most sources. The region G159.6−18.5 is easily detected with
WISE, although it has very weak radio continuum emission in
the GB6 survey Gregory et al. (1996). Andersson et al. (2000)
also detected weak radio continuum emission from this nebula.
There is no single radio continuum survey covering the entire
sky that can be used for this work. Instead, we use a variety
of radio continuum surveys, listed in Table 1. Because of the
different spatial scales probed by these surveys, it is useful to
examine all available radio data, even if these data cover the
same Galactic zone. This is especially true in the first Galactic
quadrant where there are numerous high-quality surveys. For
example, we find that the VGPS is the most sensitive survey
extant for extended diffuse emission and is useful for identifying
large diffuse H ii regions, but MAGPIS 20 cm data (Helfand
et al. 2006) boasts a higher angular resolution and point source
sensitivity than the VGPS and is useful for identifying ultra-
compact H ii regions. We encounter a similar situation in the
southern hemisphere with the SGPS and SUMSS data sets. The
simultaneous use of both high- and low-resolution radio data
produces a more complete catalog.
We examine by eye WISE and radio continuum images
spanning the entire Galactic plane, |b| 8
◦
. As in past work
(Anderson et al. 2011) we search for radio continuum emission
spatially coincident with objects having characteristic MIR
morphologies. Each field was searched at least three times
by one of us (L.D.A.), in addition to searches by other group
members. One of us (L.D.A.) determined if an object identified
by one or more group members would be included in the WISE
catalog. This method of searching each field multiple times
ensures a more complete catalog of H ii regions and H ii region
candidates. For each identified source we store in the catalog
the position and approximate radius of a circular aperture that
encloses the associated MIR emission. For complicated regions
of the Galaxy, we verify that the WISE sources are distinct using
the Spitzer GLIMPSE (Benjamin et al. 2003; Churchwell et al.
2009) and MIPSGAL (Carey et al. 2009) legacy data, when
possible.
Figure 3 shows our procedure for a 3
◦
× 1
◦
portion of the
Galaxy centered at (, b) = (30
◦
,0
◦
). The circle sizes in this
figure approximate the extent of the MIR emission associated
with each source. It is this size that is cited in the catalog.
In the inset panels, we label regions observed in RRL emis-
sion. In the left inset, G031.050+00.480, G030.956+00.599, and
G030.951+00.541 were observed in RRL emission; the other
three H ii regions in the field are “grouped” H ii region candi-
dates that are positionally associated with the three known re-
gions. The procedure of associating H ii regions and candidates
with one another to create groups is described in Section 3.2.
The middle panel shows the well-known G29 H ii region com-
plex. Because of the high density of sources in this part of the
Galaxy, however, we do not associate the H ii region candi-
date (cyan) toward the southwest of the inset with G29. The
right inset shows two H ii regions observed in RRLs: a bright
compact region (G028.983−00.603), and a more diffuse region
(G029.094−00.713). There is a compact grouped H ii region
candidate positionally associated with G028.983−00.603. To
the north is an example of an extended radio quiet candidate,
and to the south is an example of a small radio quiet source.
In addition, we perform an automated search for H ii region
candidates by matching NVSS 20 cm continuum data (Condon
et al. 1998) and MAGPIS 20 cm continuum data (Helfand
et al. 2006) with WISE point sources. To reduce the number
of spurious matches, we only include point sources that have
WISE colors [F
12
/F
22
] > 0.5 (Anderson et al. 2012b, their
Table 1) and F
12
15, where F
12
and F
22
are the 12 μm
and 22 μm WISE fluxes, respectively. This effort only yielded
another 20 H ii region candidates. This suggests that the visual
search alone is sufficient to identify most H ii regions and H ii
region candidates in the Galaxy. Unfortunately, these WISE
color criteria cannot reliably distinguish between H ii regions
and planetary nebulae (PNe). The automated search therefore
identified hundreds of PNe candidates, identifiable by their
extended latitude distribution and lack of MIR nebulosity. The
NVSS does not cover the southern sky but given the marginal
results in the northern sky automated search, we did not attempt
to repeat the automated search in the south.
Ex post facto we correlate all H ii region candidates lacking
radio continuum emission in the surveys of Table 1 with the
radio continuum observations of Urquhart et al. (2007a, 2009)
at 6 cm, Sanchez-Monge et al. (2013) at 3.6 and 1.4 cm,
and the “CORNISH” 6 cm VLA survey (Hoare et al. 2012;
Purcell et al. 2013). For Urquhart et al. (2007a, 2009) and
Sanchez-Monge et al. (2013), we search within radii of 30
and
3
