0810.3685/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: %% Modified 2004 August 30
3: %%
4: %% manuscript produces a one-column, double-spaced document:
5: %%\documentclass[12pt,manuscript]{aastex}
6: %% preprint2 produces a double-column, single-spaced document:
7: %%\documentclass[preprint2]{aastex}
8: %% Sometimes a paper's abstract is too  long to fit on the
9: %% title page in preprint2 mode. When that is the case,
10: %% use the longabstract style option.
11: %% \documentclass[preprint2,longabstract]{aastex}
12: %%
13: %%          Author-defined commands
14: %----------------------------------------------------------------------
15: %
16: %  modified to include rtr's superior LaTeXmanship
17: %
18: \def\tmbcom#1{{\bf TMB: #1}}
19: \def\ldacom#1{{\bf LDA: #1}}
20: %
21: % draft 7: lda 29 March 2006
22: % draft 8: tmb 
23: % draft 9: lda June/July 2007
24: % draft10: tmb August 2007
25: % final draft: tmb 29 Aug 2007
26: %
27: % ----- define organization
28: \newcommand{\nraoblurb}{The National Radio Astronomy Observatory is
29: a facility of the National Science Foundation operated under cooperative
30: agreement by Associated Universities, Inc.}
31: % ----- units
32: \newcommand{\m}{$\,{\rm m}$}
33: \newcommand{\cm}{$\,{\rm cm}$}
34: \newcommand{\kpc}{$\,{\rm kpc}$}
35: \newcommand{\K}{\ensuremath{\,{\rm K}}}
36: \newcommand{\ks}{${\,{\rm km\, sec^{-1}}}$}
37: \newcommand{\msun}{$\,M_\odot$}
38: \newcommand{\gyr}{$\,{\rm Gyr}$}
39: \newcommand{\mhz}{$\,{\rm MHz}$}
40: \newcommand{\ghz}{$\,{\rm GHz}$}
41: \newcommand{\percc}{$\,{\rm cm^{-3}}$}
42: \newcommand{\hr}{$\,{\rm hr}$}
43: \newcommand{\kms}{\ensuremath{\,{\rm km\,s}^{-1}}}
44: \newcommand{\degree}{\,^\circ}
45: \newcommand{\degper}{\ensuremath{\rlap.{^{\circ}}}}
46: \newcommand{\arcmper}{\rlap.{^{\prime}}}
47: \newcommand{\arcsper}{\ensuremath{{\rlap.{^{\prime\prime}}}}}
48: \newcommand{\jy}{\ensuremath{{\rm \,Jy}}}
49: \newcommand{\jyb}{$\rm \,Jy\,beam^{-1}$}
50: \newcommand{\mjy}{\,mJy}
51: \newcommand{\mjyb}{$\rm \,mJy\,beam^{-1}$}
52: \newcommand{\microjy}{\,$\mu$Jy}
53: \newcommand{\microjyb}{$\rm \,\mu Jy\,beam^{-1}$}
54: % ----- atoms etc.
55: \newcommand{\h}[1]{$^{#1}{\rm H}$}
56: \newcommand{\hi}{{\rm H\,}{{\sc i}}}
57: \newcommand{\hii}{{\rm H\,}{{\sc ii}}}       
58: \newcommand{\he}[1]{$^{#1}{\rm He}$}
59: \newcommand{\heiii}{{\rm He\,}{{\sc iii}}}
60: \newcommand{\li}[1]{$^{#1}{\rm Li}$}
61: \newcommand{\car}[1]{$^{#1}{\rm C}$}
62: \newcommand{\ci}{{\rm C\,}{{\sc i}}}
63: \newcommand{\cii}{{\rm C\,}{{\sc ii}}}
64: %
65: % 12 and 13 CO
66: \newcommand{\co}{\ensuremath{^{12}{\rm CO}}}
67: \newcommand{\cor}{\ensuremath{^{13}{\rm CO}}}
68: \newcommand{\wco}{\ensuremath{W_{\rm CO}}}
69: %
70: %  he-3 commands
71: \newcommand{\hethree}{\ensuremath{{}^3{\rm He}}}
72: \newcommand{\hefour}{\ensuremath{{}^4{\rm He}}}
73: \newcommand{\hepthree}{\ensuremath{{}^3{\rm He}^+}}
74: \newcommand{\hep}[1]{$^{#1}{\rm He}^{+}$}
75: \newcommand{\hepr}[1]{$^{#1}{\rm He}^{+}/{\rm H}^{+}$}
76: \newcommand{\hepp}[1]{$^{#1}{\rm He}^{++}$}
77: \newcommand{\heppr}[1]{$^{#1}{\rm He}^{++}/{\rm H}^{+}$}
78: \newcommand{\her}[1]{$^{#1}{\rm He}/{\rm H}$}
79: \newcommand{\lir}[1]{$^{#1}{\rm Li}/{\rm H}$}
80: %
81: % recombination lines
82: \newcommand{\hal}{${\rm H}\,91\alpha$}
83: \newcommand{\heal}{${\rm He}\,91\alpha$}
84: \newcommand{\heps}{${\rm H}\,154\epsilon$}
85: \newcommand{\heta}{${\rm H}\,171\eta$}
86: \newcommand{\halp}{${\rm H}\,\alpha$}
87: \newcommand{\halpha}{\ensuremath{\langle{}\,{\rm H}{}\,\rangle{}\,\alpha}}
88: % ----- misc
89: \newcommand{\nexpo}[2]{\ensuremath{{#1}\times 10^{#2}}}
90: \newcommand{\expo}[1]{\ensuremath{10^{#1}}}
91: \newcommand{\ngc}[1]{NGC\thinspace #1}
92: \newcommand{\jon}[1]{J\thinspace #1}
93: \newcommand{\etab}{$\eta_{\rm \,b}$}
94: \newcommand{\vdag}{(v)^\dagger}
95: \newcommand{\tex}{$T_{\rm ex}$}
96: %----------------------------------------------------------------------
97: %
98: \slugcomment{For ApJSuppl: final draft --- lda, \today}
99: 
100: \shorttitle{CO in Galactic H~{\small \bf II} Regions}
101: \shortauthors{Anderson et al.}
102: 
103: \begin{document}
104: 
105: \title{The Molecular Properties of Galactic H\,{\small \bf II} Regions}
106: 
107: \author{L. D. Anderson\altaffilmark{1}, T. M. Bania\altaffilmark{1}, 
108: J. M. Jackson\altaffilmark{1}, D. P. Clemens\altaffilmark{1}, 
109: M. Heyer\altaffilmark{2}, R. Simon\altaffilmark{3}, R. Y. Shah\altaffilmark{4},
110: \& J. M. Rathborne\altaffilmark{5}}
111: 
112: \altaffiltext{1}{Institute for Astrophysical Research,
113: 725 Commonwealth Ave., Boston University, Boston MA 02215, USA.}
114: \altaffiltext{2}{Department of Astronomy, University of Massachusetts,
115: Amherst, MA 01003, USA.}
116: \altaffiltext{3}{Physikalisches Institut, Universit\"{a}t zu K\"{o}ln,
117: 50937 Cologne, Germany}
118: \altaffiltext{4}{MIT Lincoln Laboratory, 244 Wood Street, Lexington, MA 02420}
119: \altaffiltext{5}{Harvard-Smithsonian Center for Astrophysics, 60 Garden
120: St., Cambridge, MA 02138, USA}
121: 
122: \begin{abstract}
123: 
124: We derive the molecular properties for a sample of 301 Galactic \hii\
125: regions including 123 ultra compact (UC), 105 compact, and 73 diffuse
126: nebulae. We analyze all sources within the BU-FCRAO Galactic Ring
127: Survey (GRS) of \cor\ emission known to be \hii\ regions based upon
128: the presence of radio continuum and \cm-wavelength radio recombination
129: line emission.  Unlike all previous large area coverage \cor\ surveys,
130: the GRS is fully sampled in angle and yet covers $\sim 75$ square
131: degrees of the Inner Galaxy.  The angular resolution of the GRS
132: (46\arcsec ) allows us to associate molecular gas with \hii\ regions
133: without ambiguity and to investigate the physical properties of this
134: molecular gas.  We find clear CO/\hii\ morphological associations in
135: position and velocity for $\sim 80\%$ of the nebular sample.  Compact
136: \hii\ region molecular gas clouds are on average larger than UC
137: clouds: $2 \farcm 2$ compared to $1 \farcm 7$.  Compact and UC \hii\
138: regions have very similar molecular properties, with $\sim 5 ~{\rm K}$
139: line intensities and $\sim 4 \kms$ line widths.  The diffuse \hii\
140: region molecular gas has lower line intensities, $\sim 3 ~{\rm K}$,
141: and smaller line widths, $\sim 3.5 \kms$.  These latter characteristics
142: are similar to those found for quiescent molecular clouds in the GRS.
143: Our sample nebulae thus show evidence for an evolutionary sequence
144: wherein small, dense molecular gas clumps associated with UC \hii\
145: regions grow into older compact nebulae and finally fragment and
146: dissipate into large, diffuse nebulae.
147: 
148: \end{abstract}
149: 
150: \keywords{\hii\  regions --- ISM: abundances, clouds, atoms, evolution, 
151: lines, and bands, structure --- radio lines: ISM}
152: 
153: \section{INTRODUCTION\label{sec:intro}}
154: 
155: In the classical view of \hii\ region formation OB stars form inside Giant
156: Molecular Clouds (GMCs).  Any newly formed OB star emits extreme
157: ultraviolet (EUV) radiation ($> 13.6$ eV) and ionizes the surrounding
158: medium of a molecular cloud, creating an \hii\ region. The ionizing
159: photons have more energy than is necessary to ionize the gas, however,
160: and thus some excess energy heats the ambient gas. Because of the
161: pressure difference between the cold, natal molecular cloud ($\sim$\,30
162: K) and the \hii\ region ($\sim$\,\expo{4}\,K), the ionization front
163: expands into the molecular cloud.
164: 
165: A photodissociation region (PDR) exists beyond the ionization
166: front. The PDR is a zone where photons of energies lower than 13.6 eV
167: ionize elements with low ionization potentials and dissociate
168: molecules \citep[see][ for a review of dense PDRs]{hollenbach97}.
169: Within the PDR there are three important boundaries. At $A_V$ $\simeq
170: 1$, a dissociation front exists where $\rm H_2$ replaces H as the
171: dominant species. Further from the emitting star, at $A_V \simeq 4$,
172: there is a boundary between $\rm C^{+}/ \rm C/ \rm CO$ where carbon
173: becomes stored predominantly in the CO molecule. The final boundary of
174: the PDR occurs at the $\rm O/ \rm O_2$ transition, $A_V \simeq
175: 10$. The PDR is therefore a transition region between the \hii\ region
176: and the molecular cloud; it is completely ionized at one boundary and
177: completely molecular at the other. Since all \hii\ regions should
178: produce PDRs, in this simplified view all \hii\ regions should have
179: associated molecular gas when they first form.  As the nebulae evolve,
180: however, the OB stars can travel far enough to leave the environs of
181: their natal molecular clouds.
182: 
183: Complicating this simple scenario is the fact that molecular clouds are
184: clumpy and inhomogeneous on all scales \citep[e.g.,][]{falgorone96,
185: kramer98, simon01}.  In a clumpy medium, EUV photons can penetrate to
186: different depths, creating non-spherical ionization
187: fronts. Additionally, \hii\ regions evolve by moving away from their
188: natal cloud environments and displacing the local gas. Understanding the
189: complicated interaction between young stars, \hii\ regions, and
190: molecular gas is crucial to the study of massive star formation and the
191: impact massive stars have on their environment.
192: 
193: The molecular component of \hii\ regions has been studied in detail by
194: many authors. Most studies, however, have focused mainly on
195: ultra-compact (UC) \hii\ regions \citep[e.g.,][]{churchwell90, kim03}.
196: When compact \hii\ regions were observed in CO, they were often only
197: observed with single pointings using single-dish telescopes
198: \citep[e.g.,][]{brand84, whiteoak82, russeil04}. As our results show,
199: CO gas is frequently offset from the nominal position of an \hii\
200: region. A completely sampled map is required to understand fully the
201: dynamics and properties of molecular gas associated with \hii\
202: regions. Furthermore, a sample with \hii\ regions at all evolutionary
203: stages is necessary to understand how this interaction progresses as
204: the \hii\ region ages.
205: 
206: Here we describe a large scale study of the molecular properties of
207: diffuse, compact, and ultra-compact Galactic \hii\ regions using fully
208: sampled CO maps.  We trace these molecular properties using the \cor\,
209: $J= 1\rightarrow 0$ emission mapped by the Boston University--Five
210: College Radio Astronomy Observatory (hereafter BU--FCRAO) Galactic
211: Ring Survey
212: \citep[GRS: ][]{jackson06}.  CO is an excellent tracer of molecular
213: material because of its high abundance and the fact that it is
214: rotationally excited at densities common in molecular clouds
215: ($\ga$\,500\percc).
216: 
217: \section{Milky Way Surveys\label{surveys}}
218: 
219: \subsection{Galactic H~{\bf\small II} Regions\label{catalog}}
220: % Lockman sample
221: 
222: Our Galactic \hii\ region source sample is compiled from the
223: \hii\ region radio recombination line (RRL) catalog of \citet{lockman89}
224: (hereafter L89), which consists of nearly 500 RRL observations with a
225: 3\arcmin\ beam at positions of known continuum emission. These continuum
226: emission sources were drawn from the 5 \ghz\, continuum survey of
227: \citet{altenhoff79}.  All positions in \citet{altenhoff79} with a peak
228: flux density greater than 1 \jyb\ and many sources with peak flux
229: densities down to 0.8 \jyb\ were observed, unless the source was in a
230: very confused region or was a known supernova remnant (SNR).
231: %
232: Since \citet{altenhoff79} has coverage from $l=357\fdg 5$ to $60\degr$
233: and $\vert b\vert < 1\degr$, the L89 catalog is complete down to a flux
234: limit of at least 1 \jy\ at 5 \ghz\ in this region of the Galaxy. This
235: sky coverage entirely overlaps and is much larger than the GRS survey
236: zone (discussed in \S \ref{GRS}), so the L89 catalog is complete at
237: least down to 1\jyb\ over the extent of the GRS.
238: %
239: The sources in L89 are the classical \hii\ regions first detected in
240: the 1970's at cm-wavelengths using RRLs.  Although they are often
241: misidentified, these are not the ultra compact, high density \hii\
242: regions that are being studied mostly using radio interferometers.
243: Below, to distinguish these classical \hii\ regions from UC nebulae,
244: we will call the L89 nebulae ``compact'' \hii\ regions.
245: 
246: Our nebular sample also includes diffuse \hii\ regions from
247: \citet{lockman96} (hereafter L96). This catalog consists of RRL
248: measurements at 6\cm\ ($\sim 6\arcmin$ beam) and 9\cm\ ($\sim 9\arcmin$
249: beam) toward 130 faint, extended continuum sources.  Most of these
250: sources are drawn from \citet{altenhoff79} and have peak flux densities
251: greater than 0.5 \jyb.  The L89 survey is actually the pilot study of
252: this diffuse sample; it contains 40 diffuse sources.
253: 
254: Recently a large catalog of 1442 Galactic \hii\ regions was compiled
255: from 24 published studies of Galactic \hii\ regions
256: \citep{paladini03}. These \hii\ regions were found using single-dish,
257: medium resolution (few arcminute beamwidths) observations. For our
258: purposes, however, this catalog is not useful because nebular positions
259: are given to an accuracy of only 6\arcmin.
260: 
261: Finally, our source sample contains UC \hii\ regions taken from
262: \citet{wc89a}, \citet{kurtz94}, \citet{watson03}, and
263: \citet{sewilo04}. The \citet{wc89a} UC sources were selected by: (1) the
264: presence of a small or unresolved radio source, (2) a spectrum
265: consistent with free-free emission, and (3) strong FIR emission. The UC
266: sources from \citet{kurtz94}, \citet{watson03}, and \citet{sewilo04}
267: were selected based on the \citet{wc89b} criteria: {\it IRAS} flux
268: density ratios log$(F_{25}/F_{12}) \ge 0.57$, log$(F_{60}/F_{12}) \ge
269: 1.30$, and $F_{100} \ge 1000 \jy$ ($\ge 700 \jy$ for \citet{watson03}),
270: where $F_{\lambda}$ is the {\it IRAS} flux density at $\lambda\,
271: \micron$.
272: %here%
273: \subsection{{\bf The $^{\bf 13}$}CO Galactic Ring Survey\label{GRS}}
274: 
275: We use the BU--FCRAO \cor\, Galactic Ring Survey data\footnote[1]{Data
276: available at http://www.bu.edu/galacticring/} \citep{jackson06} to
277: characterize the molecular properties of all \hii\ regions in the GRS. The GRS traces the 5 \kpc\ molecular ring discovered by
278: \citet{burton75} and \citet{ss75}.  This annulus of enhanced CO emission
279: dominates the inner Galaxy's structure and harbors most of the Galaxy's
280: star formation regions. The GRS sky coverage spans $18 \degr < l < 55
281: \degr$ and $\vert b\vert < 1 \degr$. Additional, incomplete sky coverage
282: is available for $14 \degr < l < 18 \degr$ over the same latitude range.
283: The GRS covers a total of 74 square degrees. The GRS maps the
284: distribution of emission from the $J = 1 \rightarrow 0$ ($\nu_0 = 110.2$
285: \ghz) rotational transition of \cor.  The \cor\ isotopologue is $\sim 50$ times
286: less abundant than \co\, and hence has a much smaller optical
287: depth. This decreased optical depth yields smaller line widths and
288: gives a cleaner separation of individual velocity components along any
289: specific line of sight compared to previous \co\ surveys.
290: 
291: The GRS, with a spectral resolution of $0.21 \kms$, an angular
292: resolution of $46\,\arcsec$, and a $22 \, \arcsec$ angular sampling,
293: improves upon all previous large scale CO surveys.  It is the only fully
294: sampled (in solid angle) large scale \cor\ survey extant. The GRS
295: improves upon the Bell Labs \cor\, survey \citep{lee01} that has a
296: spectral resolution of $0.68 \kms$, an angular resolution of $103
297: \arcsec$ and $180 \arcsec$ angular sampling.  The GRS also has better
298: resolution than previous \co\, surveys. For example, the University of
299: Massachusetts Stony Brook survey \citep{sanders86} has a spectral
300: resolution of $1.0 \kms$, an angular resolution of $45 \arcsec$ and
301: $180 \arcsec$ angular sampling. The Columbia/CfA \co\, survey
302: \citep{dame01} has a spectral resolution of $0.18 \kms$, an angular
303: resolution of 450\arcsec\, and $225\arcsec$ to $450 \arcsec$ angular
304: sampling. Compared to GRS, {\it all} of these surveys are severely
305: undersampled in angle. The GRS maps have spectra observed at positions
306: separated by $\sim \frac{1}{2}$ the telescope's beamwidth (0.48 HPBW,
307: actually). This, together with high spectral resolution allows us to
308: separate individual \cor\, components cleanly. Thus we can for the
309: first time study the molecular properties of Galactic \hii\ regions
310: free from angular sampling bias.
311: 
312: \section{H~{\bf\small II} Region Source Sample\label{sample}}
313: 
314: Here we study the molecular properties of all known \hii\ regions in the
315: zone mapped by the GRS. Our sample of 301 \hii\ regions contains 123 UC,
316: 105 compact, and 73 diffuse nebulae.  For UC nebulae without RRL
317: measurements in the original papers, we compile RRL velocities from
318: \citet{afflerbach96} and \citet{araya02} or, if the UC and compact
319: sources were co-spatial, from L89.  The compact nebulae are from the L89
320: catalog. The majority of our diffuse regions are from L96, but a small
321: number are from the pilot survey of diffuse regions in L89.  This final
322: nebular sample results from further vetting of the \S 2 sources using
323: data from several recently completed Galactic scale sky surveys that
324: overlap the GRS zone.
325: 
326: % vetting process
327: We verify the existence, classification, and position of each \hii\
328: region by examining the radio continuum and infrared emission at its
329: nominal position. For the radio continuum emission, we primarily use
330: the 21\,cm VLA Galactic Plane survey
331: \citep[VGPS:][]{stil06}. In addition to the 21\,cm
332: \hi\ line emission data cubes, the VGPS generated 21cm continuum images
333: over the range $18\arcdeg < l < 67\arcdeg, |b| < 1-2$ with $\sim 1
334: \arcmin$ resolution.  In addition to the VLA measurements, the VGPS 
335: used a Green Bank Telescope 21\cm\ survey to provide the zero spacing data.  
336: The VGPS is therefore sensitive to both large and small scale emission.  We also use the 20\,cm data
337: from the Multi-Array Galactic Plane Imaging Survey
338: \citep[MAGPIS:][]{helfand06}. These data were collected with the VLA
339: operating in B-, C-, and D-configurations and have a resolution of
340: $\sim 6 \arcsec$.  The 20 \cm\ MAGPIS data cover the range $5\arcdeg <
341: l < 48\fdg 5, |b| < 0\fdg8$.  We find MAGPIS to be best suited for
342: verifying UC nebulae, while the VGPS is better suited for verifying
343: compact and diffuse \hii\ regions.  For the infrared emission, we use
344: the 8\,\micron\ data from the Galactic Legacy Infrared Mid-Plane
345: Survey Extraordinaire
346: \citep[GLIMPSE:][]{benjamin03} and 24\,\micron\ data from the MIPS Inner
347: Galactic Plane Survey (MIPSGAL: Carey et al. 2008 in preperation).
348: Both infrared surveys have coverage beyond the extent of the GRS.
349: %\ldacom{This reference: (Broadbent, A., Osborne, J. L., \& Haslam,
350: %C. G. T. 1989, \mnras, 237, 381) shows that FIR (60 um) emission is correlated
351: %with thermal radio sources}
352: 
353: % Continuum peak
354: For inclusion in our sample, we require a continuum peak at the
355: position of each \hii\ region. The UC regions were identified by their
356: IRAS colors and the compact and diffuse regions were located in radio
357: continuum maps with $2 \farcm 6$ resolution. Both of these
358: identification methods have some level of error that can be reduced
359: through correlation with high resolution radio continuum data. For UC
360: regions, continuum observations are necessary to confirm that the
361: nebula is an \hii\ region, and not a dense protostellar clump. The
362: sources from \citet{wc89a} and \citet{kurtz94} were confirmed to be UC
363: regions with VLA continuum observations, but the majority of sources
364: from \citet{watson03} and \citet{sewilo04} have not yet been confirmed
365: with high resolution ratio data. We exclude 11 UC sources, 2 compact
366: sources and 2 diffuse sources that do not have significant continuum
367: emission.
368: 
369: % SNR, IR
370: We remove all known SNR from our sample by comparing our positions
371: with the catalog of \citet{green06}\footnote[2]{Available at
372: http://www.mrao.cam.ac.uk/surveys/snrs/)} as well as with two recent
373: catalogs of SNRs by \citet{brogan06} and \citet{helfand06}. Both of
374: these recent catalogs compute the spectral index of SNR candidates using
375: 20\,\cm\ and 90\,\cm\ VLA continuum data and rely on the anticorrelation
376: of SNRs with infrared emission (8\,\micron\ {\it MSX} data in the case of
377: \citet{brogan06} and 21\,\micron\ {\it MSX} data for \citet{helfand06}). We
378: exclude 13 SNRs from our sample that were found in these catalogs, as
379: well as one found in \citet{gaensler99}. There are a comparable number
380: of sources that are spatially coincident with SNRs, but which have a
381: strong IR component. We believe these are \hii\ regions in locations
382: that have produced multiple generations of stars.  These sources are
383: retained in our sample.  We also remove an additional six sources that
384: do not have infrared emission since they most likely are non-thermal.
385: 
386: Finally, we determine if the classification (UC, compact, diffuse) and
387: position of each source are correct, and remove duplicate sources. We
388: require UC \hii\ regions to have small ($\lesssim 1\arcmin$) bright knots of continuum
389: emission, compact regions to be larger bright continuum sources, and
390: diffuse regions to have faint extended continuum emission. There are
391: many cases where an UC region was mistakenly identified as a compact or
392: diffuse region in the L89 and L96 catalogs. This misidentification is
393: due to the fact that UCs are unresolved with the 2\farcm6 beam of
394: \citet{altenhoff79} from which L89 and L96 drew their positions. We
395: exclude 69 compact and diffuse nebulae whose positions are coincident
396: with UC regions.  Many of the UC \hii\ regions found in the GRS are
397: in large complexes with individual UC components separated by angular
398: distances less than the $46\arcsec$ GRS beam. We treat these
399: complexes as single UC regions because they share a common molecular gas
400: clump at the GRS resolution.
401: 
402: Table \ref{tab:anomalous} lists the 38 \hii\ regions we cull from our
403: sample because of the criteria just described.  Table
404: \ref{tab:anomalous} gives the source name, the reason for exclusion from
405: our sample, and the reference if the source is a known SNR. Source names
406: identify the nebular type: UC (``U''), compact (``C''), or diffuse
407: (``D'') . This source name convention will be followed throughout this
408: paper.
409: %
410: Figure \ref{fig:lv} shows the longitude-velocity position of our
411: nebular sample. The symbols in Figure \ref{fig:lv} indicate the
412: nebular type: UC (small filled circles), compact (medium filled
413: circles), or diffuse (large open circles).  Table \ref{tab:diffuse}
414: gives the properties of the nebulae in our sample.  Listed are the
415: source name, its position in Galactic and equatorial coordinates and
416: its RRL velocity\footnote[3]{The RRL velocities here are in the
417: kinematic Local Standard of Rest (LSR) frame using the radio
418: definition of the Doppler shift.  The kinematic LSR is defined by a
419: solar motion of 20.0 \kms\ toward ($\alpha, \delta$) = (18$^{\rm h},
420: +30\arcdeg$)[1900.0].} with its $1 \sigma$ error.  Altogether our 301
421: \hii\ region sources probe 266 unique directions since 33 nebulae have
422: RRL emission at several different velocities that presumably
423: originates from physically distinct nebulae located along the line of
424: sight.  For some \hii\ regions, we change the classification based on
425: the morphology of the VGPS and MAGPIS radio continuum emission.  We
426: also change the position of a few \hii\ regions based on this
427: continuum emission if the position is obviously incorrect.  All such
428: changes are noted with footnotes in Table \ref{tab:diffuse}.
429: 
430: \section{Finding Molecular Gas Associated with Galactic H~{\bf\small II} Regions\label{sec:ID}}
431: 
432: To characterize the properties of molecular gas associated with Galactic
433: \hii\ regions we must establish reliable morphological correlations in
434: ({\it l, b, V\/})--space between the \cor\ gas and the nebulae. The
435: GRS is a large ($\sim10 ~{\rm Gbyte}$), dataset that contains very
436: complex \cor\ emission line structure.  The ($l, b, V_{\rm LSR}$)
437: structure of the molecular gas near the nebulae can be very
438: complicated.  Any given line of sight frequently contains multiple
439: emission lines.  There probably is no single algorithm that can
440: uniquely give reliable CO/\hii\ regoim ({\it l, b, V\/}) morphological
441: correlations. We therefore use a suite of software tools to analyze
442: the GRS data cubes in a variety of ways.
443: 
444: \subsection{Analysis Software \label{sec:software}}
445: 
446: To maximize the power and flexibility of our analysis we wrote a large
447: suite of IDL procedures for spectral analysis rather than use any of the
448: standard single-dish radio astronomy software packages.  We used the
449: single-dish radio astronomy TMBIDL\footnote[4]{Written by T. M. Bania
450: and available at http://www.bu.edu/iar/research/dapsdr/} software as our
451: starting point.  TMBIDL was originally written to analyze NRAO Green
452: Bank Telescope data. (TMBIDL was the inspiration that led to the NRAO
453: GBTIDL\footnote[5]{See http://gbtidl.sourceforge.net/} software.)  The
454: TMBIDL software emulates and improves upon many of the features of the
455: NRAO UniPops analysis program.  It includes Gaussian and polynomial line
456: fitting, data visualization, data manipulation, etc.  The TMBIDL code
457: can easily be modified to analyze data from any single dish radio
458: telescope.
459: 
460: We wrote additional IDL procedures to interface and analyze GRS data
461: within the TMBIDL environment. TMBIDL was created to analyze single
462: spectra, so we added ({\it l-V\/}), ({\it b-V\/}), and ({\it l-b}) mapping
463: tools to better visualize the GRS ($l, b, V$) data cubes.  The basic
464: visualization is a normalized contour map of the CO emission.  Using
465: these tools, we found that the morphology of the CO emission at the
466: positions of the \hii\ regions was often very complex.
467: 
468: To gain further control over the visualization of the GRS data, 
469: we wrote GUI-based software\footnote[6]{Available for download at
470: http://people.bu.edu/andersld/}
471: %
472: to analyze images extracted from the GRS ({\it l,b,V\/}) FITS data
473: cubes.  This software provides powerful GUI tools to extract, image, and
474: analyze subcubes for each sample \hii\ region.  For our analysis we
475: imaged various quantities for an ({\it l,b}) zone surrounding each
476: nebula.  The user can, for example, easily modify the velocity channel
477: whose \cor\ line intensity is being imaged over the mapped region.  One
478: can also quickly create and display integrated intensity CO maps, \wco\
479: [K \kms], as well as arbitrarily vary the velocity range of the
480: integration, $\Delta V$.  (This is done with a kernel based algorithm
481: that is fast and efficient.)  Sub-images and regions can be created for
482: any image, saved and then reloaded at any time. The GUI uses 
483: DS9\footnote[7]{Available for download at
484: http://hea-www.harvard.edu/RD/ds9/} syntax to define regions.   
485: %
486: These regions can have all the basic DS9 shapes.  We added a
487: ``threshold'' region that selects all contiguous pixels above a
488: user-defined threshold level that surround a given image pixel.  Using
489: this thresholding tool one can identify and analyze arbitrarily
490: complex morphologies.
491: 
492: This GUI-software is fully integrated with TMBIDL.  The user can, for
493: example, export the spectra within any region to TMBIDL for spectral
494: analysis.  The ability to scan quickly through velocity channels, create
495: integrated intensity images, select pixels and fit Gaussians to spectra
496: --- all within a single application --- is extremely powerful.
497: 
498: \subsection{Correlation of Molecular Gas and H~{\bf\small II} Regions}
499: 
500: Our goal is to find a morphological coincidence in ({\it l,b,V\/}) --
501: space between the CO gas and the \hii\ region. After a coincidence is
502: established, we want to characterize the molecular gas using spectral
503: fits to the \cor\ emission. The UC positions are in general known to
504: an accuracy greater than the $22 \arcsec$ GRS pixel spacing. For the
505: compact and diffuse \hii\ regions, our positions are accurate to a few
506: arcminutes.  The RRL LSR velocities are accurate to $\sim\, 0.1 \kms$
507: (from Gaussian fits).  Although the ({\it l,b,V\/}) position of each nebula is
508: accurately known, establishing a robust set of criteria for
509: identifying a real molecular/\hii\ physical association is nontrivial.
510: 
511: The CO emission maps of \hii\ regions can be quite complex due to GMC
512: structure and PDR/ionization front interactions.  For example, one
513: expects the molecular and ionized gas velocities to diverge as an \hii\
514: region evolves.  Once an OB star forms within a GMC its ionization front
515: (IF) expands rapidly at first, reaching the nebular Str\"{o}mgren radius
516: in $\lesssim 10^5$ yr.  The IF pushes the surrounding GMC molecular gas
517: outwards.  \citet{dyson97} show that as the IF expands, it rapidly slows
518: until it is expanding at $\sim 10 \kms$ when it reaches the
519: Str\"{o}mgren radius.  Since \hii\ regions are generally older than
520: $10^5$ yr (with the possible exception of some UC \hii\ regions), the
521: maximum difference between the RRL and the associated molecular gas
522: should be $\lesssim 10 \kms$.
523: 
524: Our nebular sample has \hii\ regions of different ages and thus should
525: show evolutionary effects.  Diffuse \hii\ regions should be older than the UC
526: nebulae, and thus have had more time to evolve away from and displace their natal
527: clouds. The molecular gas in diffuse \hii\ regions should show a weaker
528: association or may not be present at all. We expect the majority of UC
529: and compact \hii\ regions to be associated with a molecular clump
530: \citep[see, e.g.,][]{kim03}. Because of these complications, our
531: CO/\hii\ analysis is comprised of a series of distinct investigations.
532: 
533: \subsubsection{Single Position Spectrum Analysis}\label{sec:singlepointing}
534: 
535: We first examine the GRS \cor\ spectrum at the nominal position of each
536: \hii\ region. Most previous studies of the molecular component of \hii\ regions
537: were made using single pointings, e.g., \citet{whiteoak82}; Russiel \&
538: Castets (2004).  All report that the majority of \hii\ regions have
539: associated CO.  \citet{whiteoak82} found in their survey of \co\
540: emission from Southern \hii\ regions that molecular gas within 5 \kms\
541: of the RRL velocity had large line intensities, and therefore was
542: probably associated with the \hii\ region. In their analysis of Southern
543: compact \hii\ regions using both \co\ and \cor, \citet{russeil04} argued
544: that 10 \kms\ is a better criterion for determining a molecular/\hii\
545: association.
546: 
547: To make a single pointing CO/\hii\ comparison we calculate an average
548: \cor\ spectrum at the position of each \hii\ region by convolving the
549: GRS datacube (which is oversampled in angle) with the FCRAO telescope
550: beam (HPBW = $46 \arcsec$).  We then search this average spectrum for
551: a \cor\ emission line peak at the \hii\ region RRL velocity.
552: Specifically, we look for emission above 0.5\K\ brightness
553: temperature, $T_{\rm MB}$, and within $\pm 2.5 \kms$ of the \hii\
554: region LSR velocity.
555: 
556: Our brightness temperature limit is chosen to be well above the GRS
557: noise.  The GRS data have a typical RMS sensitivity of $\sigma({\rm
558: T_{MB}})=0.27\K$.  
559: %Because the GRS data were taken using two methods,
560: %position switching and OTF mapping, the error in the beam-convolved
561: %average spectrum does not always decrease as much as theoretically
562: %expected since some spectra share common off positions.  
563: The beam convolved average spectrum has a factor of $\sim\,2$ decrease
564: in noise compared to a single GRS position.  Thus our 0.5\K\
565: search criterion is a $\sim3\sigma$ limit.
566: 
567: Using the 0.5\,K and $\pm 2.5\kms$ criteria, only 52 \% of the nebular
568: sample shows a CO/\hii\ association.  Repeating this procedure with the
569: same intensity requirement, but with velocity ranges of $\pm 5.0 \kms$
570: and $\pm 7.5 \kms$, we find that, respectively, $70\%$ and $79\%$ of
571: \hii\ regions meet these criteria.  Certainly increasing the velocity
572: range further still will yield a greater number of nebulae matching the
573: association criteria, but relaxing the association definition in this
574: way also increases the chance of a misidentification and the possibility
575: of blending multiple velocity components.
576: %We therefore henceforth adopt $\pm 7.5 \kms$ as our velocity criterion 
577: %for molecular gas association.
578: 
579: \subsubsection{\wco\, Integrated Intensity Map Analysis}\label{sec:map_analysis}
580: 
581: We use the GRS data to make an ({\it l, b\/}) contour map of the \cor\
582: integrated intensity, \wco\ (K \kms), in order to provide information
583: about the spatial distribution of the molecular gas.  The main
584: weakness of the single pointing method of searching for CO/\hii\
585: region associations is that sources with molecular gas offset from the
586: nominal position of the \hii\ region are not counted as detections.
587: For each nebula we use TMBIDL to make normalized \wco\ contour maps
588: that are ${\it l} \times {\it b} = 12 \arcmin \times 12 \arcmin$
589: ($\sim\,30 \times 30$ GRS pixels) in size.  The map \wco\ is
590: calculated by integrating each spectrum over $15 \kms$ centered at the RRL velocity.  The map peak \wco\
591: is used to normalize the \wco\ value for each pixel.  The final
592: product is a normalized contour map for each nebula in the sample.
593: 
594: We then search each map for \wco\, peaks and note the distance from
595: the 80\% peak \wco\ contour to the nominal \hii\ region position.  We
596: define any source where this distance lies within the \hii\ region
597: positional error bars to be a positive detection and any source where
598: this distance is just outside of the error bars (roughly twice the
599: positional uncertainty) to be an ambiguous detection.  All sources not
600: meeting these positional criteria are deemed to be non-detections. We
601: find that 70\% of our sample nebulae show positive detections, 14\%
602: have ambiguous detections, and 16\% show no correlation between the
603: \hii\ region position and the \cor\ emission.  Somewhat suprisingly,
604: adding the molecular spatial distribution information to the CO/\hii\
605: association criterion did not add significantly to the detection rate.
606: 
607: \subsubsection{CO ($\/l,b,V$) Data Cube Analysis Procedure}\label{sec:gui}
608: 
609: Clearly, the molecular emission in the GRS is complicated and
610: difficult to characterize.  These experiments in establishing a
611: CO/\hii\ region association demonstrate the need for a more
612: sophisticated analysis. We use the \S \ref{sec:software} GUI software
613: to search the GRS \cor\, data cubes in ($l, b, V$) parameter space.
614: For each nebula we make a series of ($l,b$) \wco\ images and search
615: for CO/\hii\ region associations.  We follow the four step iterative
616: procedure described below and illustrated in Figure \ref{fig:3plot}.
617: 
618: % intitial search of pos-pos maps
619: \begin{enumerate}
620: \item  We first find the velocity range of the molecular emission 
621: associated with the \hii\ region. We examine single velocity channel
622: images (position-position ({\it l, b\/}) maps) centered at the nominal
623: position of the \hii\ region at its RRL velocity with overlaid VGPS
624: 21cm continuum contours.  We scan through these single channel images
625: over $\pm 10 \kms$ of the source RRL velocity, searching for the
626: channel where the molecular emission at the position of the \hii\
627: region has the highest intensity. If we are able to identify a
628: molecular clump near the position of the \hii\ region, we extract the
629: spectra from the voxels near this molecular clump.  We fit a Gaussian
630: to the (unweighted) average spectrum of this extracted emission and
631: record the line center and FWHM of the emission line.  Frequently, the
632: molecular emission at the \hii\ region position is either absent or
633: has a morphology that is difficult to characterize from single channel
634: maps. For $\sim\,25\%$ of our nebulae we are unable to make a
635: molecular gas association from these single channel maps.
636: 
637: %creation of int-int
638: \item Next we make an integrated intensity map, \wco, by summing the
639: intensities at a given ({\it l,b}) over the range of velocities found in
640: step (1). If the source has an unambiguous CO/\hii\ association, we
641: calculate \wco\, over the velocity range $V \pm \Delta V/2$, where $V$
642: is the step (1) Gaussian line velocity and $\Delta V$ is its FWHM line
643: width.  If the source has an ambiguous step (1) association, we
644: calculate \wco\, centered at the source RRL velocity over the range $\pm
645: 10 \kms$.
646: 
647: % pixel selection
648: \item We then use the $W_{\rm CO}$ image created in step (2) to
649: find pixels with molecular emission associated with the \hii\ region.
650: We find the brightest emission near the \hii\ region in the \wco\,
651: image and select all contiguous pixels that have values above a
652: threshold determined independently for each \hii\ region.  The
653: threshold is varied until a small number of pixels are selected ---
654: typically 20 to 30.  The exact number is set by the molecular clump's
655: intensity profile.  Small clumps with a sharply peaked intensity
656: profile have fewer pixels, whereas larger clumps with a ``plateau'' of
657: emission have more pixels.  This method seeks to isolate single
658: clumps in order to preserve the line peak intensity and minimize
659: blending of multiple velocity components.
660: 
661: % fitting
662: \item All the GRS spectra within the ($l,b$) region selected in step (3)
663: are used to calculate an average \cor\, spectrum.  We then fit the
664: fewest possible number of Gaussian components to the spectrum in order
665: to maximize the intensity of the peak CO emission.  The majority
666: ($\sim\,80\%$) of our sources are adequately fit with only one
667: Gaussian component. Many nebulae, however, do not have clean
668: Gaussians profiles, but rather show structure that often suggests a fainter,
669: wider line superposed on a brighter, narrower line.  
670: \end{enumerate}
671: 
672: Our analysis procedure is summarized in Figure \ref{fig:3plot} for the
673: \hii\ region U43.24$-$0.05. Panels A and B depict step 1. Panel A
674: shows a single channel GRS map at the velocity nearest the RRL
675: velocity where the molecular clump has the highest intensity.  The
676: cross marks the nominal \hii\ region position.  For clarity we have
677: not shown the VGPS continuum emission contours.  The black box shows
678: the ($l,b$) positions of the voxels from which we extract spectra to
679: produce the average spectrum shown in Panel B. The vertical line in
680: Panels B and D marks the RRL velocity.  Using the velocity range of
681: the emission line shown as dashed lines in Panel B, we create the
682: integrated intensity image shown in Panel C (step 2). This image has
683: been smoothed with a $3 \times 3$ Gaussian kernel.  The black outline in Panel C shows contiguous
684: integrated intensity values above a threshold (step 3).  This region
685: and its threshhold is defined by visual inspection of the image.  This
686: is the \cor\ emission we deem to be associated with the \hii\
687: region.  Finally, we extract the spectra from these ($l,b$) positions
688: to produce the average spectrum shown in Panel D (step 4).  For this
689: source, our method separates the two emission lines that are blended
690: in the panel B spectrum and preserves the peak line intensity.
691: 
692: % pos-pos vs int-int
693: % TMB is not convinced that this adds anything.
694: % LDA: let's cut it - I don't like it anymore.
695: %The properties of the molecular gas associated with this \hii\ region
696: %are derived from Gaussian fits to the Panel D average spectrum. This
697: %spectrum shows significant non-Gaussian line wings.  If instead we
698: %defined the association zone by selecting pixels from single channel
699: %position-position images, we would find much cleaner lines. Since our
700: %goal was to characterize the properties of all the molecular gas
701: %associated with an \hii\ region, and not simply to have clean fits to
702: %spectral lines, we chose to select pixels from the integrated
703: %intensity images at the expense of cleaner fits.
704: 
705: % completeness of method amazing misuse of the term 'completeness'
706: In principle, we could iterate our analysis to locate and fit the
707: molecular emission more accurately.  Using the Panel D Gaussian fit, we
708: could create a new integrated intensity image, define a new CO/\hii\
709: association region, and fit a Gaussian to this new average spectrum.  We
710: did this for 10 test cases and found only only minimal changes that were
711: not significantly different from a single pass analysis. 
712: 
713: % benefits of procedure
714: This procedure has many advantages.  We are able to characterize the
715: spectral properties of molecular gas distributions that have arbitrary
716: morphologies.  We minimize our assumptions at every stage of the
717: process.  By first examining the ({\it l,b}) images of each source at
718: individual velocity channels, we limit false detections that may arise
719: from integrated intensity images containing multiple velocity
720: components blended together.  By extracting the spectra from regions
721: defined in the integrated intensity images, we make no assumptions
722: based on the visual appearance of the molecular emission at individual
723: velocity channels.  By using a variable threshold to select pixels, we
724: are able to characterize molecular structures with arbitrary
725: morphology.  This threshold definition ensures clean spectral fitting;
726: the spectra are not contaminated by adjacent pixels that would lower
727: the line intensity and might increase the width of the fitted Gaussian
728: line.  Finally, since we analyze an average of many spectra, the lines
729: we fit have a much greater signal to noise ratio than a single
730: pointing spectrum.  The Gaussian fit uncertainties in the spectral
731: line parameters we derive are thus minimized.
732: 
733: % ellipse fits
734: After we establish a CO/\hii\ region association, we characterize the
735: angular size of the molecular cloud by fitting an ellipse to the \wco\
736: spatial distribution.  Since the association defined in Panel C
737: of Figure \ref{fig:3plot} is a \wco\ threshold value that is
738: determined independently for each source, the size of an ellipse
739: fitted to this zone would have little physical meaning.  We therefore
740: define a new threshold selected region that is uniform for
741: the entire nebular sample.  This new region is defined by a threshold
742: set to 80\% of the \wco\ peak inside the original association zone.
743: The fitting algorithm calculates the semimajor and semiminor axes
744: using the ``mass density'' of pixel locations: the clustering of pixel
745: locations along the {\it l} and {\it b} directions.  The fitted
746: ellipse to this region is thus a uniform estimate of the size of the
747: molecular gas associated with the \hii\ region.
748: 
749: The properties we derive for the CO/\hii\ region associations are
750: summarized in Table \ref{tab:properties}.  For each nebula we list the
751: parameters of the ellipse and Gaussian line fits. All errors are $1
752: \sigma$ uncertainties.  Given are the ellipse centroid in Galactic
753: coordinates, its size, semi-major, $a$, and semi-minor, $b$, axes,
754: together with the position angle measured from north toward increasing
755: Galactic longitude.  The ellipse size is the geometric mean diameter,
756: $2\sqrt{a b}$.  If there are multiple spectral components in the
757: Gaussian fit only the properties of the brightest are listed.  We also
758: use only this brightest component in our subsequent analysis.  The
759: fitted line parameters given are the center velocity, intensity (in
760: main beam brightness temperature units), and FWHM line width, $\Delta
761: V$. Table \ref{tab:properties} also lists some additional nebular
762: properties derived below in \S \ref{sec:discussion}: the CO excitation
763: temperature, the \cor\ column density, and the nebular confidence
764: parameter, $CP$.
765: 
766: \section{Discussion \label{sec:discussion}}
767: 
768: Our analysis here provides a large sample of molecular cloud/\hii\
769: region associations whose physical properties are well characterized.
770: In \S\ref{sec:ID} we show that the molecular emission is often
771: morphologically complex and offset from the nominal \hii\ region
772: position and RRL velocity.  We find that the traditional single
773: pointing analysis does not reliably detect the molecular components of
774: \hii\ regions.  It is very difficult to make a CO/\hii\ association with any
775: confidence without an image produced from a $(l, b, V)$ data cube that
776: spans the entire \hii\ region/PDR/molecular cloud interaction region.
777: 
778: %\subsection{Confidence Parameter \label{sec:confidence}}
779: 
780: To distinguish sources with an unambiguous molecular gas component at
781: the \hii\ region position and velocity from those with less robust
782: molecular gas associations, we assign a confidence parameter, $CP$, to
783: each source ranging from A to E.  Our qualitative criteria for the
784: confidence parameter are as follows: 
785: %
786: A: no ambiguity in position or velocity, the molecular gas coincident 
787: with radio continuum or shows clear signs of interaction; 
788: %
789: B: either offset somewhat in position or velocity, or in a
790: complex region of molecular gas; 
791: %
792: C: either offset in position or velocity or in a complex region, 
793: but fainter than a B source and offset further; 
794: %
795: D: diffuse emission near the correct position and velocity, but 
796: uncharacterizable due to low intensity or ambiguous morphology; 
797: %
798: E: nothing at all apparent in position and velocity. 
799: 
800: Figures \ref{fig:images_a}, \ref{fig:images_b} and
801: \ref{fig:images_cde} give representative examples of our confidence
802: parameter classification.  These $20 \arcmin \times 20 \arcmin$ \cor\,
803: integrated intensity images span the range of molecular gas
804: morphologies surrounding our sample of \hii\ regions.  The images are
805: grouped by their confidence parameter classification: Figure
806: \ref{fig:images_a} shows $CP$ A sources, Figure \ref{fig:images_b}
807: shows $CP$ B sources and Figure \ref{fig:images_cde} shows $CP$ C, D,
808: and E sources.  Because the GRS is oversampled ($22\arcsec$ pixels at
809: $46\arcsec$ HPBW), we increase the signal to noise by smoothing the
810: images with a $3 \times 3$ Gaussian filter.
811: 
812: At the top of each image we list the \hii\ region name, the \cor\ line
813: velocity (\kms) and FWHM line width (\kms), $\Delta V$, and the source's confidence
814: parameter.  These Gaussian fitted line parameters are used to
815: calculate the source's \cor\ integrated intensity, \wco (K \kms), for
816: the velocity range $V \pm \Delta V /2$.  At the bottom of each image
817: we give the \wco\, calculated from the fitted line: $W_{\rm CO} = 1.06
818: \, T_{\rm MB} \, \Delta v$, which is accurate for Gaussian line
819: shapes.  Using the procedure described in \citet{simon01}, we also
820: give an estimate of the ${\rm H_2}$ column density, $N({\rm H_2})
821: [{\rm cm^{-2}}]$.
822: 
823: The grey scale image shows the \wco\ distribution.  Normalized
824: contours are drawn at 83\%, 67\%, 50\%, 33\% and 16\% of this maximum.
825: (One can get quantitive \wco\ values using the scale bar at the
826: right.)  For \hii\ regions where we could not associate molecular gas,
827: the contour lines are dashed rather than solid and, of course, no line
828: parameters are given.
829: 
830: A bold cross marks the nominal position of the \hii\ region.  Plotted
831: in bold is the fitted ellipse described in \S \ref{sec:gui}.  If there
832: are other \hii\ regions in the field, they are marked with thinner
833: crosses. The cross arm lengths correspond to the beam size used to
834: make the measurement of the \hii\ region ($3 \arcmin$ for L89;
835: $9\arcmin$ for L96), except for the UC regions where the cross arm
836: lengths are set to 1\arcmin.  The GRS beam (HPBW = $46
837: \arcsec$) is shown in the lower left corner of each
838: image. Shown in grey in these images are VGPS 21cm continuum contours.
839: Tickmarks on these contours point downhill, towards decreasing 21cm
840: emission.
841: 
842: We are able to establish a highly confident ($CP$ A and B) CO/\hii\
843: region association for 62\% of the nebulae in our sample.  Relaxing
844: the confidence criterion to $CP$ A, B and C sources gives CO/\hii\
845: associations for 84\% of our nebular sample.  Histograms of the number
846: distribution of confidence parameter values are shown in Figure
847: \ref{fig:confidence} for UC (open), compact (hatched), and diffuse
848: (gray) nebulae.  The top left panel is the stacked histogram of the
849: distribution; the top line represents the entire sample.  For example,
850: there are a total of 113 $CP$ A sources: 78 UC, 29 compact, and 6
851: diffuse \hii\ regions.  Clockwise from here are the individual
852: histograms for the UC, compact, and diffuse nebulae.  (Most subsequent
853: figures will follow this display format.)  As is clear from Figure
854: \ref{fig:confidence}, the UC sample has the greatest number of high
855: confidence CO/\hii\ region associations.  
856: 
857: % literature on lack of molecular assoc.
858: Some \hii\ regions appear to have no molecular gas associated with
859: them. Of our 301 sources, 14 (5\%) are classified as E sources; these
860: nebulae show no \cor\, emission whatsoever.  Thirty four sources
861: (11\%) are classified as D and therefore have only diffuse emission at
862: the correct position and velocity.  Five of our E sources and 15 of
863: our D sources have multiple RRL velocities along the same line of
864: sight. In these cases, one RRL velocity is probably from the \hii\
865: region of interest while the other is likely from a nearby \hii\
866: region.  Eleven of the 12 UC regions with confidence parameter values
867: of D or E have multiple velocity components.  
868: 
869: Two UC \hii\ regions are worth mentioning individually because of the
870: nature of their molecular associations.  The UC \hii\ region
871: U33.13+0.09 has a very large velocity offset between the molecular
872: material and the RRL velocity.  L89 find a RRL velocity of 93.8 \kms\,
873: and \citet{araya02} find a RRL velocity of 87.4 \kms.  Our \cor\
874: velocity of 75 \kms\ is in agreement with the CS velocity found by
875: \citet{bronfman96}.  The morphology and linewidth of the molecular
876: emission suggest that the molecular gas is associated with the \hii\
877: region.  The UC \hii\ region U21.42$-$0.54 also has a compact
878: molecular clump at the correct ({\it l, b}\/) position.  The \cor\
879: velocity of this clump, 54 \kms, is 16 \kms\ offset from its RRL
880: velocity of 70 \kms.  This source was not detected by
881: \citet{bronfman96}.  We assign a confidence parameter value of 
882: C to U21.42$-$0.54 because of the extreme velocity offset.
883: 
884: The lack of CO/\hii\ region associations in $\sim\,20\%$ of our sample
885: is not entirely unexpected.  This has been reported in the literature
886: before, although previous studies did not have datasets that were
887: fully sampled in angle as is the GRS.  \citet{blitz82} found a lack of
888: associations in $30\%$ of the Sharpless \hii\ regions studied.
889: \citet{russeil04} found no association with CO in $\sim 20
890: \%$ of their sample of Southern compact \hii\ regions.
891: \citet{churchwell90} do not detect the dense gas tracer ammonia in
892: $\sim\,30\%$ of a sample of 84 UC \hii\ regions.
893: 
894: Our large sample of \hii\ regions contains nebulae spanning a range of
895: evolutionary stages. The UC regions, being young, are more likely to
896: lie within their natal molecular clouds where the density of \cor\,
897: should be high.  As the \hii\ region evolves, it will dissipate the
898: gas.  We should see evidence for this in the compact and diffuse
899: regions.  We therefore expect UC regions to be associated with small
900: sizes, high excitation temperatures and column densities, large line
901: widths, and bright line intensities. Diffuse regions should have large
902: sizes and smaller values for the other quantities compared to the UC
903: regions. Compact regions should lie in between these two extremes.
904: 
905: The lack of molecular gas in an \hii\ region can certainly be an
906: evolutionary effect.  These nebulae may represent an older population
907: of \hii\ regions that have had time to become displaced from the gas
908: in which they formed due to a variety of mechanisms including, stellar
909: winds, ionization fronts, high stellar space velocities, etc.  One
910: observational consequence of this scenario might be a bubble
911: morphology in the molecular gas, seen as a ring in projection.
912: \citet{churchwell06} found a large population of bubbles in the
913: GLIMPSE survey \citep{benjamin03}.  Observed at mid-infrared
914: wavelengths, GLIMPSE boasts an angular resolution 10 times that of the
915: GRS and is therefore a better diagnostic tool for locating bubble
916: features.  We indeed see bubbles in \cor\, in most \hii\ regions in
917: the GRS for sources where \citet{churchwell06} also identified
918: bubbles. These regions are usually classified as D or E sources
919: because the gas has been pushed far away from the center of the
920: continuum emission. Further investigation of this topic will be the
921: subject of a future paper.
922: 
923: \subsection{Properties of Molecular Cloud/H~{\bf\small II} Region Sources 
924: \label{sec:properties}}
925: 
926: Here we focus on the 253 nebulae with positive CO/\hii\ associations,
927: the subset of our \hii\ region sample with confidence parameter values
928: of A, B or C.  The $(l,V)$ distribution of these nebulae is shown in
929: Figure \ref{fig:co_lv} for $CP$ A (large filled circles), B (small
930: filled circles), and C (small open circles) sources.  Fully 90\% of UC
931: and compact nebulae have $CP$ A, B, or C quality CO/\hii\ associations
932: whereas only 64\% of the diffuse nebulae do.  Diffuse \hii\ regions
933: are probably older on average than either UC or compact nebulae so
934: their significantly lower associate rate with molecular gas provides
935: the first hint of evolutionary effects in our sample.
936: 
937: Table \ref{tab:groups} summarizes the mean properties we derive here
938: for this sample. Listed are the mean and standard deviation ($1
939: \sigma$) for each quantity.  This information is given for the entire
940: sample and also for various subsets of it: UC, compact, and diffuse
941: \hii\ regions as well as sources of $CP$ A, B and C.
942: Table \ref{tab:groups} lists the number of \hii\ regions in each
943: category, the absolute value of the velocity difference between the
944: \cor\, molecular clump and the RRL $V_{\rm LSR}$, the line intensity (main beam
945: brightness temperature), the FWHM line width, the size of the
946: associated molecular clump, the CO excitation temperature and the
947: \cor\, column density.
948: 
949: %\subsubsection{Column Density\label{sec:column}}
950: We estimate the average \cor\ column density towards our sources  
951: using the \citet{rohlfs} analysis:
952: 
953: \begin{equation} \label{eq:13cocolumn} 
954: N(^{13} {\rm CO})[{\rm cm^{-2}}] = 2.42 \times 10^{14} \frac{T_{\rm ex} 
955: \int \tau_{13}\, dv}{1 - {\rm exp}(-5.29)\,/\,T_{\rm ex}} 
956: \end{equation}
957: 
958: \noindent where \tex\, is the excitation temperature and $\tau_{13}$
959: is the optical depth of the \cor\, line.  We assume the \cor\ emission
960: is optically thin and use the Gaussian fit \cor\ line parameters to
961: find the optical depth integral in Eq.\ref{eq:13cocolumn}.  Both
962: the optical depth and Equation \ref{eq:13cocolumn}, however, depend on
963: the excitation temperature, \tex.
964: 
965: We use the UMSB \co\ $J = 1
966: \rightarrow 0$ survey \citep{sanders86} to estimate \tex\ for each
967: source.  We assume \co\, is optically thick and use the radiative
968: transfer equation for the $J = 1 \rightarrow 0$ transition to
969: calculate the excitation temperature from the observed main beam
970: brightness temperature, $T_{\rm MB}^{12}$, of the \co\, line:
971: 
972: \begin{equation} \label{eq:tex} 
973: T_{\rm ex} = 5.5 \left/ \ln \left( 1+
974: \frac{5.5}{T_{\rm MB}^{12}+0.82} \right) \right..
975: \end{equation} 
976: 
977: \noindent Eq. \ref{eq:tex} holds as long as: (1) the \co\, and
978: \cor\, emitting gas is in LTE at the same excitation temperature; (2)
979: this gas fills the same volume without clumping; and (3) there are no
980: background continuum sources.  (For our nebulae the \hii\ region
981: continuum is subtracted when the spectral baselines are removed.)
982: 
983: For each nebula we first compute an average spectrum from the UMSB
984: survey datacube in exactly the same way as we did for the GRS data.
985: We use the same ($l,b$) positions from the identical threshold
986: selected region for this average.  We then fit Gaussians to these
987: average spectra. If we fit multiple Gaussians to the GRS data, we
988: attempt to fit the same components to the UMSB data. The spectral
989: resolution of the UMSB survey is 1\kms, so this was not always
990: possible as lines resolved in the GRS are blended in the UMSB survey.
991: %\ldacom{For 60 sources (!), the 12co intensity is less than the 13co
992: %intensity.  They are generally smaller than average ($<2$ arcmin).  I
993: %set the 12 and 13co values equal for these sources.  I could put a
994: %less than symbol in the tables, but then all the $T_{ex}$ and
995: %$N(13co)$ values are really lower limits, which they are anyway I
996: %suppose.}
997: 
998: Because the pointings of the UMSB survey are further apart than in the
999: GRS ($180 \arcsec$ compared to $22 \arcsec$), the UMSB survey
1000: underestimates the \co\ emission for the small molecular clumps found
1001: in the GRS. For a given velocity, each pixel in the UMSB survey
1002: represents $\sim 70$ pixels in the GRS. Each threshold selected region
1003: contains $\sim 20$ GRS pixels on average, so for the majority of our
1004: sources we use only one UMSB pointing to estimate the excitation
1005: temperature appropriate for the \co\ emission.  Furthermore, this
1006: pointing can be as far away as $\sim 120\arcsec$ from the GRS
1007: position.
1008: 
1009: The distribution of excitation temperatures we derive is shown in
1010: Figure \ref{fig:t_ex}.  All \hii\ regions in our sample have very
1011: similar excitation temperatures near the standard 10\,K value assumed
1012: for molecular clouds.  We expected the excitation temperature of the
1013: UC regions in particular to be higher than this standard value as the
1014: CO gas is nearer to the exciting star. That we do not see hotter
1015: temperatures associated with younger regions is probably due to the
1016: effect of the undersampling of the \co\ emission by the UMSB survey.
1017: The mean \co\ to \cor\ ratio for molecular clumps smaller than the
1018: beam spacing of the UMSB survey, $3\arcmin$, is $\sim2$ while this
1019: ratio is $\sim 3$ for molecular clumps larger than $3\arcmin$.  Our
1020: excitation temperatures, and hence column densities, are therefore
1021: lower limits.  Because of their small size, UCs are affected more by
1022: the difference in sampling between the GRS and the UMSB survey.  The
1023: UCs also suffer from beam dilution which will lower the inferred
1024: excitation temperature.
1025: 
1026: Nevertheless, we use these excitation temperatures to compute the
1027: \cor\, column density for each source using Eq. \ref{eq:13cocolumn}.
1028: These column densities are listed in Table \ref{tab:diffuse}.  We then
1029: estimate the $\rm{H}_2$ column density,
1030: 
1031: \begin{equation} \label{eq:h2column} 
1032: N({\rm H_2}) = \left[\frac{^{12} 
1033: {\rm CO}}{^{13} {\rm CO}} \right] \times \left[\frac{\rm H_2}{^{12} {\rm
1034: CO}} \right] \times N(^{13} {\rm CO}), 
1035: \end{equation} 
1036: 
1037: \noindent by assuming constant values for these abundance ratios.
1038: Following \citet{simon01} we adopt a \co/\cor\ ratio of 45 and a $\rm
1039: H_2$/\co\ ratio of \nexpo{8}{-5}.  Our \cor\, and $\rm H_2$ column
1040: density results are summarized in Figure \ref{fig:column}.  As
1041: expected, the UC nebulae have on average the highest column densities
1042: and the diffuse nebulae the lowest.  Too, the $CP$ A sources have on
1043: average over twice the column density of the $CP$ C nebulae.
1044: 
1045: We find that the CO gas has on average only a small velocity offset
1046: from the \hii\ region RRL velocity. Figure \ref{fig:vel_diff_by_code}
1047: shows the difference between the velocity of the CO gas and the RRL
1048: velocity.  There is no difference in velocity offset between the
1049: various types of \hii\ regions.  The Gaussian fit to the entire
1050: distribution is centered at 0.4 \kms\ with a FWHM of 8.5 \kms,
1051: whereas the mean of the distribution is $0.2\pm 3.8$ \kms. The fact
1052: that the distribution is centered at zero velocity offset is to be
1053: expected for a RRL selected sample of \hii\ regions and an optically
1054: thin tracer such as \cor.
1055: 
1056: This result is in contrast to optically selected samples where a 
1057: positive velocity offset was found \citep[e.g., ][]{fich90}.  Optical
1058: samples choose specific CO/\hii\ region line of sight geometries.
1059: Face-on or edge-on blister sources such as the Orion nebula and M17
1060: should dominate these samples.  In our sample there is no preferred
1061: radial location for \hii\ regions within molecular clouds.  Because we
1062: do not know the CO/\hii\ geometry with respect to the line of sight,
1063: the absolute value of the CO/RRL velocity difference will be a direct
1064: measure of any systematic velocity offset between the molecular and
1065: ionized gas.  Table \ref{tab:groups} therefore lists the mean absolute
1066: value of the CO/RRL velocity difference; it shows that there is a
1067: $\sim\,3 \kms$ average flow velocity between the molecular and ionized
1068: gas for the nebulae in our sample.  This value is independent of the
1069: type of \hii\ region, but increases slightly as the $CP$ value decreases from A to C.
1070: 
1071: % Line Intensities
1072: Figure \ref{fig:intensity} shows the distribution of \cor\ line
1073: intensities for our sample. Based on the evolutionary model of \hii\
1074: regions the natal cloud is gradually dissipated by photo-dissociation,
1075: photo-ionization, and expanding motions.  We expect the \cor\ density
1076: to decrease as the region progresses from UC to compact and then,
1077: finally to diffuse.  Assuming \cor\, is optically thin, or at least
1078: marginally so, the higher densities found in UC regions would lead to
1079: higher line intensities, whereas compact sources should show lower
1080: line intensities, and diffuse sources the lowest. This hypothesis is
1081: only partially borne out as UC and compact regions share the same
1082: distribution, averaging $5.21 \pm 0.23 \K$ and $4.96 \pm 0.24 \K$,
1083: respectively. Diffuse regions do show lower line intensities of $3.32
1084: \pm 0.19$. The errors quoted here are the standard errors of the 
1085: mean, s.e.m. $\equiv \sigma/\sqrt{N}$.
1086: 
1087: % line widths
1088: Figure \ref{fig:linewidth} shows the distribution of \cor\ line widths
1089: for our sample. The Gaussian fit to this distribution is centered at
1090: 4.0 \kms\ with a FWHM of 3.2 \kms. The mean of this distribution is
1091: $4.20\pm0.09$ \kms. We expected UC \hii\ regions to have significantly
1092: broader lines than compact \hii\ regions because the molecular gas is
1093: closer to the exciting star and the outflows should be stronger.  Once
1094: again, this is not the case: the UC and compact distributions are very
1095: similar, averaging $4.43\pm0.13\kms$ (s.e.m) and $4.23 \pm 0.14\kms$
1096: (s.e.m.), respectively.  UC and compact regions do have broader lines
1097: than the diffuse regions which average $3.56\pm0.20 \kms$ (s.e.m.).
1098: This suggests that the central star(s) may no longer be significantly
1099: heating molecular gas near diffuse \hii\ regions. 
1100: 
1101: This distribution of line widths is comparable to that found by
1102: \citet{russeil04} in their single pointing survey of southern \hii\
1103: regions. They find that the \cor\, $J = 1 \rightarrow 0$ line has an
1104: average line width of 3.7 \kms\ with a standard deviation of 1.9
1105: \kms.  In a study of UC \hii\ regions, \citet{kim03} find an average line
1106: width of 6.8 \kms\ in the \cor\, $J=1 \rightarrow 0$ transition.  Their
1107: calculation of line width, however, was based on the average spectrum
1108: over the entire map area, which was as large as $30 \arcmin \times 40
1109: \arcmin$. For the 8 UC regions in our sample that \citet{kim03} study,  
1110: we measure a line width of 5.0 \kms\ whereas they measure 7.2
1111: \kms.  Using the same large areas to compute the line widths for these
1112: sources, we find an average line width of 7.3 \kms.
1113: 
1114: % Sizes
1115: %\subsubsection{CO Angular Sizes \label{sec:sizes}}
1116: The angular size distribution of the \hii\/CO sources is shown in
1117: Figure \ref{fig:angsize}. These sizes are defined as the geometric
1118: mean of the major ($2 \times a$) and minor ($2 \times b$) axes of the
1119: fitted ellipse, $2\sqrt{ab}$. (See \S\ref{sec:singlepointing} for our
1120: ellipse fitting procedure.)  There are 6 sources that have sizes
1121: greater than $10 \arcmin$ that are not plotted in Figure
1122: \ref{fig:angsize}. These sources are invariably clumps in a
1123: large region of extended molecular emission, which makes our method of
1124: determining the angular size unreliable.
1125: 
1126: The molecular clumps associated with UC regions do show the smallest
1127: sizes, as expected, averaging $1 \farcm 7 \pm 0 \farcm 1$ (s.e.m.). Compact
1128: \hii\ regions are slightly larger, averaging $2 \farcm 2 \pm 0 \farcm
1129: 1$ (s.e.m.).  The average size of the molecular gas associated with diffuse
1130: \hii\ regions lies in between that of UC and compact \hii\ regions,
1131: averaging $1 \farcm 9 \pm 0 \farcm 1$ (s.e.m.).  We have removed the 6
1132: sources with sizes greater than $10\arcmin$ from the statistical
1133: analysis. The molecular gas around many diffuse regions is fragmented,
1134: which leads to the small angular sizes we measure.  
1135: %In the case of a
1136: %few diffuse regions (e.g. D24.14+0.43), the star forming region is
1137: %going through a new round of star formation, and the molecular
1138: %material is therefore associated with young UC \hii\ regions or
1139: %prestellar cores.  
1140: Since we only associate a single molecular clump of contiguous pixels
1141: with each nebula, for diffuse \hii\ regions we probably have not
1142: characterized all the associated molecular gas.
1143: 
1144: %One really want to know, of course, the physical size of molecular
1145: %clumps associated with the \hii\ regions.  This requires knowing the
1146: %distances to our nebulae.  The vast majority of our sources are not
1147: %optically visible, so only kinematic distances can be gotten.
1148: %Unfortunately, all of these first Galactic quadrant sources suffer
1149: %from the kinematic distance ambiguity (KDA); their LSR velocities can
1150: %be produced by nebulae at two different distances, ``near and far'',
1151: %along the line of sight.  Resolving the KDA for our \hii\ region
1152: %sample is beyond the scope of this effort.  The KDA is resolved for
1153: %this sample by \citet{anderson08}.
1154: 
1155: \subsection{Comparison with GRS Molecular Clumps}\label{sec:GRS}
1156: 
1157: The properties of molecular clumps in the GRS were analyzed down to
1158: size scales of $\sim 1\arcmin$ (Rathborne et al. 2008, in
1159: preparation).  The contiguous pixel finding algorithm {\it CLUMPFIND}
1160: \citep{williams94} was used to locate Giant Molecular Clouds (GMCs)
1161: within the GRS.  Then, by altering the size threshold in {\it
1162: CLUMPFIND}, the clumps within the GMCs were identified and
1163: characterized.  The distribution of peak intensities and line widths
1164: for these clumps shows a Gaussian core with an exponential tail at
1165: high values of each parameter.  The break points where the
1166: distributions turn over from being dominated by the Gaussian core to
1167: being dominated by the exponential tail are roughly 4 K and 2 \kms.
1168: By number the vast majority of these GRS molecular clumps have line
1169: intensities below 4 K and line widths below 2 \kms.
1170: 
1171: The molecular gas associated with \hii\ regions has on average a
1172: greater line intensity and larger line width compared to molecular
1173: clumps in the GRS.  We plot in Figure \ref{fig:t_vs_fwhm} the line
1174: intensity verses the FWHM line width for the molecular clumps
1175: associated with our \hii\ regions: UC (filled triangles), Compact (filled circles),
1176: and Diffuse (open squares) nebulae. The solid lines divide the plot
1177: into quadrants according to the break points of the GRS clumps.  A
1178: similar plot was used by \citet{clemens88} to show that the small
1179: clouds in their optically selected molecular cloud sample were cool
1180: and quiescent.
1181: 
1182: The lower left quadrant in Figure \ref{fig:t_vs_fwhm} should be
1183: populated by cold quiescent clouds.  These objects are neither making
1184: stars nor being externally heated.  The upper left quadrant should
1185: contain a population of clumps that are heated externally.  These
1186: clouds are warm (or have high column densities), but do not have the
1187: non-thermal motions that would be present if they posessed a central
1188: star.  The upper right quadrant should contain molecular gas
1189: associated with embedded massive stars.  The lower right quadrant
1190: should contain embedded protostars.  These large line width objects
1191: are likely active sites of star formation, or near an active site.
1192: Thus the lower, $<4\K$, part of the plot has a pre-stellar population
1193: whereas the upper part contains clouds that are affected by local
1194: massive stars.
1195: 
1196: The UC and compact nebulae occupy the same region of Figure
1197: \ref{fig:t_vs_fwhm}; they have similar molecular properties.  The
1198: diffuse regions, however, have lower line intensities and line widths;
1199: they are similar to the general population of molecular clumps. These
1200: clumps are no longer being heated significantly by the ionizing star.
1201: 
1202: Most of our sources with associated \cor, 54\%, lie in the upper right
1203: quadrant of Figure \ref{fig:t_vs_fwhm} where large line intensities
1204: and broad line widths suggest active star formation.  The vast
1205: majority of GRS clumps, as well as most of the molecular clouds in
1206: \citet{clemens88}, reside in the lower left quadrant.  The bulk of our
1207: remaining nebulae, 42\%, lie in the lower right quadrant.  The UC,
1208: compact, and diffuse regions all have a significant population in this
1209: quadrant.  It is tempting to think that these low line intensities are
1210: due to the decreased optical depth of \cor\, but Figure
1211: \ref{fig:t_vs_fwhm} looks very similar to the same plot in
1212: \citet{russeil04} made using the optically thick \co\, $J=2
1213: \rightarrow 1$ transition.  
1214: 
1215: \subsection{Are Molecular Cloud/H~{\bf\small II} Region Associations Real?
1216: \label{sec:reality}}
1217: 
1218: Are these CO/\hii\ region associations really sources that are having
1219: a direct physical interaction between the \hii\ region and ambient
1220: molecular gas?  The associations are based on morphological matches in
1221: position and velocity between the \cor\ gas, RRL velocity, and radio
1222: continuum emission.  But correlation does not imply causality: these
1223: matches could in principle be a coincidental juxtaposition projected
1224: on the sky into the same solid angle by molecular clouds and \hii\
1225: regions located at entirely different places along the line of sight.
1226: 
1227: That we require a morphological match in $(l, b, V)$ -- space places a
1228: severe constraint on a false positive association.  Mere $(l, b)$
1229: coincidence is not enough; the velocity also needs to match.  The
1230: kinematic distance ambiguity in the Inner Galaxy makes it possible for
1231: the \hii\ region and CO cloud to be at different line of sight
1232: positions despite having nearly identical radial velocities.  But
1233: these are special places because only they share the same LSR
1234: velocity.  Assessing the quantitative probability of a false positive
1235: association is beyond the scope of this paper.  To our knowledge no
1236: one has yet done the detailed modelling this would require.  One needs
1237: to know the Galactic distribution of the clouds which posits a
1238: detailed knowledge of Galactic structure.  For a false positive
1239: association we require that there not be a cloud at the \hii\ region
1240: position, but that there be a cloud at the other kinematic distance.
1241: One thus needs to evaluate separately for each \hii\ region the line
1242: of sight distance derivative of the LSR velocity, $dV/dr$ in order to
1243: assess the path lengths at the near and far kinematic distances that
1244: must be populated.  This requires a detailed knowledge of Galactic
1245: kinematics, including streaming motions caused by spiral arms.  With
1246: this information one might be able to estimate the probability of a
1247: false positive association.  We probably do not know enough about
1248: either Galactic structure or Galactic kinematics to do this.
1249: 
1250: %
1251: % TMB does not buy this argument
1252: %
1253: %The typical GRS \cor\ spectrum has $\sim\,6$ discrete spectral
1254: %components.  GRS spectra sample first quandrant lines of sight that
1255: %typically span $\sim\,100\kms$.  Our \cor\ clouds have $\sim\,4\kms$
1256: %FWHM line widths. This gives $\sim\,25$ velocity bins and $6/25 \sim
1257: %25\%$ of the bins have a cloud in them. So for a random cloud
1258: %distribution the chance that any bin has a cloud is 1/4 and the chance
1259: %that it does not have a cloud is 3/4. For a false positive association
1260: %we require that there not be a cloud at the \hii\ region position, but
1261: %that there be a cloud at the other kinematic distance.  This requires
1262: %a joint probability of $1/4 \times 3/4 \sim 20\%$.  The fact that
1263: %$\sim\,20\%$ of the \hii\ regions do not have associated CO gas (\S
1264: %\ref{sec:properties} is evidence that chance line of sight
1265: %superpositions in $(l, b, V)$ -- space are rare.  A negative
1266: %association requires that both the near and far kinematic distances be
1267: %devoid of a cloud, a situation with a joint probability of 56\%.
1268: 
1269: The fact, however, that $\sim\,20\%$ of the \hii\ regions do not have
1270: associated CO gas (\S \ref{sec:properties}) is evidence that suggests
1271: chance line of sight superpositions in $(l,b,V)$ -- space are rare.
1272: Furthermore, Figure \ref{fig:t_vs_fwhm} provides strong support for
1273: the physical reality of our CO/\hii\ region associations.  Our \cor\
1274: clouds are not only near to the \hii\ regions in $(l,b,V)$ -- space,
1275: but their spectra also have the trademarks of star formation: bright
1276: lines and large line widths.  This is in marked contrast to the
1277: spectral line properties of the vast majority of GRS clouds (see \S
1278: \ref{sec:GRS}).  Only 1/3 of the GRS clumps have peak intensities $> 4
1279: \K$ whereas 55\% of our CO/\hii\ associated clouds do.  Only 1/6 of the
1280: GRS clouds have line widths $> 2 \kms$; nearly all of our clouds,
1281: 96\%, exceed this value.
1282: %
1283: We conclude that most of the CO/\hii\ region associations must be
1284: nebulae with real physical interactions between the molecular and
1285: ionized gas.
1286: 
1287: \section{The GRS H\,{\small \bf II} Region Catalog}\label{sec:catalog}
1288: 
1289: Our analysis here produced a catalog of Figure \ref{fig:images_a} type
1290: images and physical properties for the sample of 301 Galactic \hii\
1291: regions.  We created a
1292: website\footnote[8]{http://www.bu.edu/iar/hii\_regions} to
1293: give everyone access to this information. In addition to the images of
1294: nebular \wco, this website has the average
1295: \cor\ spectrum of each source as well as all the information found in
1296: Tables \ref{tab:diffuse} and \ref{tab:properties}.  We expect this
1297: website to be an evolving database compiling additional information
1298: about these nebulae as it becomes available.
1299: 
1300: \section{Future Work}\label{sec:future}
1301: 
1302: Our sample of CO/\hii\ region associations will enable many further
1303: studies of the properties of star forming regions at all stages of
1304: their evolution.  The most important parameter that is missing here is
1305: the distance to each nebula.  Knowing the distance would enable us to
1306: derive the intrinsic physical properties of each nebula, establishing
1307: their physical sizes and turning column densities and line intensities
1308: into masses and luminosities.
1309: 
1310: \citet{anderson08}[AB hereafter] use \hi\ absorption studies to 
1311: derive kinematic distances toward all the \hii\ regions with
1312: associated molecular gas.  All our nebulae are in the first Galactic
1313: quadrant, so their distances are degenerate due to the kinematic
1314: distance ambiguity.  Using the fact that \hi\ absorbs thermal
1315: continuum from the \hii\ region, AB use the VGPS 21cm \hi\
1316: emission line maps to remove this degeneracy \citep[see][]{kuchar94}.
1317: This is a proven technique as there is sufficient residual cold \hi\
1318: associated with almost all GRS molecular clouds to produce significant
1319: absorption \citep{jackson02, kolpak03, flynn04}.  We shall then use
1320: these distances to analyze this nebular sample and derive the physical
1321: properties of the dust and gas (ionized, atomic \& molecular).  The
1322: completion of the {\it Spitzer} GLIMPSE \citep{benjamin03} and MIPSGAL
1323: (Carey et al. 2008 in preperation) surveys, together with the GRS
1324: \citep{jackson06}, MAGPIS \citep{helfand06}, NVSS \citep{condon98} 
1325: and the VGPS surveys enable for the first time a multi-wavelength
1326: analysis of the physical properties and evolutionary state of a large
1327: sample of inner Galaxy \hii\ regions.  Due to our large sample size,
1328: we will be able to find examples of \hii\ regions at all evolutionary
1329: stages.
1330: 
1331: %We hope that multi-wavelenth analyses will allow us to distinguish
1332: %between evolutionary and geometrical effects.  The \hii\ region RRL
1333: %position and velocity targets a location where GRS \cor\ and VGPS 21
1334: %cm continuum images show the distribution and relative geometry of
1335: %molecular and ionized gas. The {\it Spitzer} GLIMPSE and MIPSGAL
1336: %images reveal the highly embedded star cluster that ionizes the
1337: %nebula, probe the distribution of the warm dust, and trace the
1338: %polycyclic aromatic hydrocarbon (PAH) emission from the PDR that lies
1339: %between the molecular and ionized zones. We shall use these high
1340: %resolution images to search for evidence of evolutionary effects such
1341: %as: (1) velocity gradients in the \cor\ GRS and \hi\ VGPS data
1342: %indicating outflows, especially in UC \hii\ regions; (2) a bubble
1343: %morphology in evolved \hii\ regions in both GLIMPSE and GRS data with
1344: %24 \micron\ MIPSGAL data possibly filling in the bubble
1345: %\citep[see][for examples of this]{churchwell06}; and (3) spatially
1346: %coincident molecular gas and dust in the UC regions with no evacuated
1347: %cavity.  The combination of high resolution radio data from the GRS
1348: %and VGPS with the GLIMPSE and MIPSGAL surveys will allow us to
1349: %associate dust features with specific velocities.
1350: 
1351: %Finally, for some nebulae we plan to obtain ground-based NIR (JHK)
1352: %spectroscopy for the brightest stars using the Mimir instrument on the
1353: %Perkins telescope at Lowell Observatory. Our ability to locate and find
1354: %the spectral types of the illuminating stars in our \hii\ regions,
1355: %together with an estimate of the energy required to displace the
1356: %molecular gas and dust would provide an estimate of an \hii\ region's
1357: %age. This type of analysis would provide information on the evolution of
1358: %the star formation rate for massive stars in our Galaxy.
1359: 
1360: \section{Summary}
1361: 
1362: We analyzed the GRS \cor\, molecular gas associated with all known \hii\
1363: regions covered by the GRS using multiple analysis techniques. Our
1364: sample includes 301 regions: 123 UC, 105 compact and 73 diffuse \hii\
1365: regions. We found that 80\% of our \hii\ regions showed positive
1366: molecular associations, with UCs having the highest association
1367: percentage and diffuse regions the lowest. About 5\% of our sample
1368: showed no molecular emission whatsoever. We hypothesize that some of
1369: these non-detections represent an older population of \hii\ regions
1370: where the molecular gas has been displaced from the central star or
1371: stars.
1372: %
1373: We found that the molecular properties of UC and compact \hii\ regions
1374: are quite similar, with line widths averaging $\sim 4 \kms$ and \cor\,
1375: column densities of about $3.5 \times 10^{16} {\rm \,cm^{-2}}$. The
1376: molecular gas associated with diffuse regions has properties more
1377: consistent with quiescent clouds. The molecular gas properties of our
1378: sample nebulae are consistent with an evolutionary sequence wherein
1379: small, dense molecular gas clumps associated with UC \hii\ regions
1380: grow into older compact nebulae and finally fragment and dissipate
1381: into large, diffuse nebulae.
1382: 
1383: \acknowledgments
1384: This publication makes use of molecular line data from the Boston
1385: University-FCRAO Galactic Ring Survey (GRS). The GRS is a joint
1386: project of Boston University and Five College Radio Astronomy
1387: Observatory, funded by the National Science Foundation under grants
1388: AST-9800334, AST-0098562, \& AST-0100793.
1389: \nraoblurb
1390: 
1391: %%%% one needs to do the *entire* bibliography in the style below before it
1392: %%%% works:  leaving out only one []{} construction screws things up....
1393: 
1394: \begin{thebibliography} {}
1395: 
1396: %% ApJ rules:  up to 3 authors, list them all; > 3, use One et al. 
1397: %%
1398: %%\bibitem[Auri\`ere(1982)] {aur82} Auri\`ere, M.  1982, \aap, 109, 301
1399: 
1400: %%\bibitem[Allen, Heaton \& Kaufman (2004)] {allen04} 
1401: %%Allen,R.J., Heaton, H.I., \& Kaufman, M.J. 2004 \apj, 608, 314
1402: 
1403: \bibitem[Altenhoff et al. (1979)] {altenhoff79} 
1404: Altenhoff, W.J., Downes, D., Pauls, T., \& Schraml, J. 1979, \aaps, 35, 23
1405: 
1406: \bibitem[Afflerbach et al. (1996)]{afflerbach96} 
1407: Afflerbach, A., Churchwell, E., Accord, J.M., Hofner, P., Kurtz, S., \&
1408: DePree, C.G. 1996, \apjs, 106, 423
1409: 
1410: \bibitem[Anderson \& Bania (2008)]{anderson08}
1411: Anderson, L.D. \& Bania, T.M. 2008, \apj, submitted
1412: 
1413: \bibitem[Araya et al. (2002)] {araya02}
1414: Araya, E., Hofner, P., Churchwell, E., \& Kurtz, S. 2002, \apjs, 138, 63
1415: 
1416: %%\bibitem[Balser et al. (1994)] {bbbrw94} 
1417: %%Balser, D.S., Bania, T.M., Brockway, C.J., Rood, R.T., \& Wilson, T.L. 
1418: %%1994, \apj, 430, 667 
1419: 
1420: %%\bibitem[Balser et al. (1997)] {bbrw97} 
1421: %%Balser, D.S., Bania, T.M., Rood, R.T., \& Wilson, T.L. 1997, \apj, 
1422: %%483, 320
1423: 
1424: %%\bibitem[Bania, Rood \& Balser (2002)] {brb02} Bania, T.M., Rood, 
1425: %%R.T., \& Balser, D.S. 2002, Nature, 415, 54
1426: 
1427: %%\bibitem[Bania, Rood \& Wilson (1987)] {brw87} 
1428: %%Bania, T.M., Rood, R.T., \& Wilson, T.L. 1987, \apj, 323, 30
1429: 
1430: %%\bibitem[Bania et al. (1997)] {bbrww97} 
1431: %%Bania, T.M., Balser, D.S., Rood, R.T., Wilson, T.L., \& Wilson, T.J. 
1432: %%1997, \apjs, 113, 353 
1433: 
1434: \bibitem[Benjamin et al. (2003)]{benjamin03}
1435: Benjamin, R. A., et al. 2003, \pasp, 115, 953
1436: 
1437: \bibitem[Blitz, Fich, \& Stark (1982)]{blitz82}
1438: Blitz, L., Fich, M., \& Stark, A.A. 1982, \apjs, 49, 183
1439: 
1440: \bibitem[Brand et al. (1984)]{brand84}
1441: Brand, J., van der Bij, M.D.P., de Vries, C.P., Leene, A., Habing, H.J., Israel, F.P., 
1442: de Graauw, T., van de Stadt, H., \& Wouterloot, J.G.A. 1984, \aap, 139, 181
1443: 
1444: \bibitem[Brand(1986)]{brand86}
1445: Brand, J. 1986, PhD Thesis, Leiden Univ. (Netherlands)
1446: 
1447: \bibitem[Brogan et al. (2006)]{brogan06}
1448: Brogan, C. L., Gelfand, J. D., Gaensler, B. M., Kassim, N. E., \& Lazio, T. J, 2006, /apj, 639, 25
1449: 
1450: \bibitem[Bronfman, Nyman, \& May (1996)]{bronfman96}
1451: Bronfman, L., Nyman, L-A. \& May, J. 1996, \aaps, 115, 81
1452: 
1453: \bibitem[Burton et al. (1975)] {burton75}
1454: Burton, W. B., Gordon, M.A., Bania, T.M., \& Lockman, F.J. 1975,
1455: \apj, 202, 30
1456: 
1457: %\bibitem[Carey et al. (2007) in preperation]{carey07}
1458: %Carey, S. J., et al. 2007, forthcoming
1459: 
1460: %%\bibitem[Chaison (1971)] {chas71}
1461: %%Chaison, E.J. 1971, \apjl, 167,61
1462: 
1463: %\bibitem[Cesaroni, et al. (1991)]{cesaroni91}
1464: %Cesaroni, R., Walmsley, C.M., K\"{o}mpe, C., \& Churchwell, E. 1991,
1465: %\aap, 252, 278
1466: 
1467: %\bibitem[Cintia, et al. (2006)]{cintia06}
1468: %Cintia, Q., Rood, R.T., Balser, D.S., \& Bania, T. M. 2006, \apjs, 165, 338
1469: 
1470: \bibitem[Churchwell, Walmsley, \& Cesaroni (1990)]{churchwell90}
1471: Churchwell, E., Walmsley, C.M., \& Cesaroni, R. 1990, \aaps, 83, 119 
1472: 
1473: %bubbles
1474: \bibitem[Churchwell et al. (2006)]{churchwell06}
1475: Churchwell, E., et al. 2006, \apj, 649, 759
1476: 
1477: \bibitem[Clemens(1985)]{clemens85}
1478: Clemens, D. P. 1985, \apj, 295, 422
1479: 
1480: \bibitem[Clemens \& Barvainis(1988)]{clemens88}
1481: Clemens, D. P. \& Barvainis, R. 1988, \apjs, 68, 257
1482: 
1483: \bibitem[Clark et al. (2003)]{clark03}
1484: Clark, J. S., Egan M. P., Crowther P. A., Mizuno D. R., Larionov V. M., \&
1485: Arkharov A. 2003, \aap, 412, 185
1486: 
1487: %\bibitem[Combes (1991)]{combes91}
1488: %Combes, F. 1991, \araa, 29, 195
1489: 
1490: \bibitem[Condon et al. (1998)]{condon98} 
1491: Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perley, R. A.,
1492: Taylor, G. B., \& Broderick, J. J. 1998, \aj, 115, 1693
1493: 
1494: %\bibitem[Crowther \& Conti (2003)]{crowther03}
1495: %Crowther, P.A. \& Conti, P.S. 2003, \mnras, 343, 143
1496: 
1497: \bibitem[Dame, Hartmann \& Thaddeus (2001)] {dame01}
1498: Dame, T. M., Hartmann, D., \& Thaddeus, P. 2001, \apj, 547, 792
1499: 
1500: %%\bibitem[Dickel, Dickel \& Wilson (1978)] {dickel78} 
1501: %%Dickel, J.R., Dickel, H.R., \& Wilson, W.J. 1978, \apj, 223, 840
1502: 
1503: %%\bibitem[Dickey \& Lockman (1990)] {dickey90} 
1504: %%Dickey, J.M., \& Lockman, F.J. 1990, \araa, 28, 215
1505: 
1506: \bibitem[Dyson \& Williams (1997)]{dyson97} 
1507: Dyson, J. E., \& Williams, D. A. 1997, The Physics of the Interstellar
1508: Medium (2nd ed.; Bristol: Inst. Physics)
1509: 
1510: %%\bibitem[Dupree (1974)] {dupree74}
1511: %%Dupree, A.K. 1974, \apj, 187,25
1512: 
1513: \bibitem[Falgarone \& Phillips (1996)]{falgorone96}
1514: Falgarone E. \& Phillips, T.G. 1996, \apj, 472, 191
1515: 
1516: \bibitem[Fich, Dahl, \& Treffers (1990)]{fich90}
1517: Fich, L., Dahl, G., \& Treffers, R. 1990, \aj, 99, 622
1518: 
1519: \bibitem[Flynn et al. (2004)]{flynn04}
1520: Flynn, E. S., Jackson, J. M., Simon, R., Shah, R. U., Bania, T. M. 
1521: \& Wolfire, M. 2004, ASP Conference Series, 317, 44 
1522: 
1523: %\bibitem[Fontani, Cesaroni, Caselli, \& Olmi (2002)]{fontani02}
1524: %Fontani, F., Cesaroni, R., Caselli, P., \& Olmi, L. 2002, \aap, 389, 603
1525: 
1526: %%\bibitem[F\"urst et al. (1990)] {frrr90} 
1527: %%F\"urst, E., Reich, W., Reich, P., \& Reif, K. 1990, \apjs, 85, 691-803
1528: 
1529: \bibitem[Gaensler, Gotthelf, \& Vasisht (1999)]{gaensler99}
1530: Gaensler, B. M., Gotthelf, E. V., \& Vasisht, G. 1999, \apj,, 526, 37
1531: 
1532: %%\bibitem[Genzel et al. (1988)] {ghjs88} 
1533: %%Genzel, R., Harris, A.I., Jaffe, D.T., \& Stutzki, J. 1988, \apj, 332, 1049
1534: 
1535: %%\bibitem[Gordon \& Churchwell (1970)] {gordon70} 
1536: %%Gordon, M.A., \& Churchwell, E.B. 1970, \aap, 9, 307
1537: 
1538: % from paladini03
1539: %\bibitem[Green (2000)] {green00}
1540: %Green, D.A., 2000, A Catalog of Galactic Supernova Remnants (2000
1541: %August version), Mullard Radio Astronomy Observatory, Cavendish
1542: %Laboratory (Cambridge, United Kingdom)
1543: 
1544: \bibitem[Green (2006)] {green06}
1545: Green, D.A. 2006, A Catalog of Galactic Supernova Remnants (2006
1546: April version), Mullard Radio Astronomy Observatory, Cavendish
1547: Laboratory (Cambridge, United Kingdom)
1548: 
1549: %\bibitem[G\'{o}mez (1995)]{gomez95}
1550: %G\'{o}mez, Y., Garay, G., \& Lizano, S. 1995, \apj, 453, 727
1551: 
1552: %%\bibitem[Haynes, Caswell \& Simons (1978)] {haynes78} 
1553: %%Haynes, R.F., Caswell, J.L., \& Simons, W.J. 1978, Aust. J. Phys. 
1554: %%Atrophys. Suppl., 45, 1
1555: 
1556: \bibitem[Helfand et al. (2006)]{helfand06}
1557: Helfand, D.J., Becker, R.H., White, R.L., Fallon, A., and Tuttle,
1558: S. 2006, \aj, 131, 2525
1559: 
1560: %\bibitem[Hofner, Wyrowski, Walmsley, \& Churchwell (2000)]{hofner00}
1561: %Hofner, P., Wyrowski, F., Walmsley, C. M., \& Churchwell, E. 2000, \apj, 536, 393
1562: 
1563: %%\bibitem[Hollenbach \& Tielens (1999)] {ht99} 
1564: %%Hollenbach,D.J., \& Tielens, A.G.G.M. 1999, Rev. Mod. Phys., 71, 1
1565: 
1566: \bibitem[Hollenbach \& Tielens (1997)] {hollenbach97}
1567: Hollenbach, D. J., \& Tielens, A. G. G. M. 1997, \araa, 35, 179
1568: 
1569: %%\bibitem[Howe et al (1991)] {howe91} 
1570: %%Howe, J.E., Jaffe, D.T., Genzel, R., \& Stacey, G. 1991, \apj, 373, 158
1571: 
1572: %% paper has 10 authors
1573: %%\bibitem[Huang et al. (1999)] {huang99} 
1574: %%Huang, M., Bania, T.M., Bolatto, A. et al. 1999, \apj, 517, 282 
1575: 
1576: %\bibitem[Simon et al.(2001)]{2001ApJ...551..747S} Simon, R., Jackson, J.~M., Clemens, D.~P., Bania, T.~M., \& Heyer, M.~H.\ 2001, \apj, 551, 747
1577: 
1578: \bibitem[Jackson et al. (2002)] {jackson02}
1579: Jackson, J. M., Bania, T. M., Simon, R., Kolpak, M., Clemens, D. P. 
1580: \& Heyer, M. 2002, \apj, 566, 81 
1581: 
1582: \bibitem[Jackson et al. (2006)] {jackson06}
1583: Jackson, J. M., Rathborne, J.M., Shah, R.Y., Simon, R., Bania, T.M.,
1584: Clemens, D.P., Chambers, E.T., Johnson, A.M., Dormody, M. \& Lavoie, R. 
1585: 2006, \apjs, 163, 145 
1586: 
1587: %\bibitem[Jaffe \& Wilson (1981)] {jaffe81} 
1588: %Jaffe, D.T., \& Wilson, T.L. 1981, \apj, 246, 113
1589: 
1590: %%\bibitem[Jenkins \& Tripp (2001)] {jt01} 
1591: %%Jenkins, E.B., \& Tripp, T.M. 2001, \apjs, 137, 297
1592: 
1593: %\bibitem[Langer \& Penzias (1990)]{langer90}
1594: %Langer, W.D. \& Penzias, A.A. 1990, \apj, 357, 477
1595: 
1596: %%\bibitem[Keene et al. (1985)] {keene85} 
1597: %%Keene, J., Blake, G.A., Phillips, T.G., Huggins, P.J., \& Beichman, 
1598: %%C.A. 1985, \apj, 299. 967
1599: 
1600: %%\bibitem[Keene et al. (1998)] {keene98} 
1601: %%Keene, J., Schilke, P., Kooi, J., Lis, D.C., Mehringer, D.M., \& 
1602: %%Phillips, T.G. 1998, \apjl, 494, L107
1603: 
1604: %\bibitem[Kim \& Koo (2001)]{kim01}
1605: % Kim, K. \& Koo, B. 2001, \apj, 549, 979
1606: 
1607: \bibitem[Kim \& Koo (2003)]{kim03}
1608: Kim, K. \& Koo, B. 2003, \apj, 596, 362
1609: 
1610: \bibitem[Kolpak et al. (2003)]{kolpak03}
1611: Kolpak, M.A., Jackson, J.M., Bania, T.M., \& Clemens, D.P. 2003, \apj, 582, 756
1612: 
1613: \bibitem[Kramer et al. (1998)]{kramer98}
1614: Kramer, C., Stutzki, J., R\"{o}hrig, R., \& Corneliussen, U. 1998, \aap, 329, 249 
1615: 
1616: \bibitem[Kuchar \& Bania (1994)]{kuchar94}
1617: Kuchar, T.A. \& Bania, T.M. 1994, \apj, 436, 117
1618: 
1619: \bibitem[Kurtz et al. (1994)] {kurtz94}
1620: Kurtz, S.,  Churchwell, E.,  Wood, D.O.S. \& Myers, P. 1994, \apjs, 91, 659
1621: 
1622: %%\bibitem[Kulesa et al. (2005)] {kulesa05} 
1623: %%Kulesa, C.A., Hungerford, A.L., Walker, C.K., Zang, X., \& Lane, A.P. 
1624: %%2005, \apj, 625, 194
1625: 
1626: %%\bibitem[Kulkarni \& Heiles (1987)] {kh87} 
1627: %%Kulkarni, S.R. \& Heiles, C. 1987, in Interstellar Processes, ed. 
1628: %%Hollenbach, D.J., \& Thronson, A. (Reidel, Dordrecht), 87
1629: 
1630: \bibitem[Kwok, Volk, \& Bidelman (1997)]{kwok97}
1631: Kwok, S., Volk, K., \& Bidelman W.P. \apjs, 1997, 112, 557
1632: 
1633: %\bibitem[L\'{e}bron, Rodr\'{i}guez \& Lizano (2001)]{lebron01}
1634: %L\'{e}bron, M.E., Rodr\'{i}guez, L.F. \& Lizano, S. 2001, \apj, 560, 806
1635: 
1636: \bibitem[Lee et al. (2001)] {lee01}
1637: Lee, Y., Stark, A. A., Kim, H., \& Moon, D. 2001, ApJS, 136, 137
1638: 
1639: %%\bibitem[Lehner, Wakker \& Savage (2004)] {lws04} 
1640: %%Lehner, N., Wakker, B.P., \& Savage, B.D. 2004, \apj, 615, 767 
1641:  
1642: \bibitem[Lockman (1989)] {lockman89} 
1643: Lockman, F. J. 1989, \apjs, 71, 469
1644: 
1645: \bibitem[Lockman, Pisano, \& Howard (1996)] {lockman96} 
1646: Lockman, F. J, Pisano, D. J., \& Howard, G. J., \apj, 472, 173
1647: 
1648: %\bibitem[Milam et al. (2005)]{milam05}
1649: %Milam, S.N., Savage, C., Brewster, M.A., \& Ziurys, L.M. 2005, \apj, 634, 1126
1650: 
1651: %%\bibitem[Natta, Walmsley \& Tielens (1994)] {nwt94} 
1652: %%Natta, A., Walmsley, C.M., \& Tielens, A.G.G.M. 1994, \apj, 428, 209
1653: 
1654: %\bibitem[Olmi, Cesaroni, \& Walmsley (1993)]{olmi93}
1655: %Olmi, L., Cesaroni, R., \& Walmsley, C. M. 1993, \aap, 276, 489
1656: 
1657: \bibitem[Paladini et al. (2003)] {paladini03}
1658: Paladini R., Burigana C., Davies R. D., Maino D., Bersanelli M.,
1659: Cappellini B., Platania P., \& Smoot G. 2003, \aap, 397, 213
1660: 
1661: %%\bibitem[Papadopoulos, Thi \& Viti (2004)] {ptv04} 
1662: %%Papadopoulos, P.P, Thi, W.-F., \& Viti,S. 2004, \mnras, 351, 147 
1663: 
1664: %\bibitem[Peimbert et al. (1992)] {Peimbertandus} 
1665: %Peimbert, M., Rodr{\'\i}guez, L.F., Bania, T.M., Rood, R.T., \& Wilson, 
1666: %T.L. 1992, \apj, 395, 484
1667: 
1668: %%\bibitem[Phillips \& Huggins (1981)] {ph81} 
1669: %%Phillips, T.G., \& Huggins,P.J. 1981, \apj, 251, 533
1670: 
1671: %%\bibitem[Reich et al. (1990)] {rfrr90} 
1672: %%Reich, W., F\"urst, E., Reich, P., \& Reif,K. 1990, \apjs, 85, 633
1673: 
1674: %%\bibitem[Rickard, Zuckerman \& Palmer (1977)] {} 
1675: %%Rickard, L.J., Zuckerman, B., \& Palmer,P. 1977, \apj, 218, 659
1676: 
1677: \bibitem[Rohlfs \& Wilson (1996)] {rohlfs}
1678: Rohlfs, K. \& Wilson, T.L. 1996, Tools of Radio Astronomy, (3rd ed.; Heidelberg: Springer)
1679: 
1680: %%\bibitem[Rood, Bania \& Wilson (1984)] {rbw84} 
1681: %%Rood, R.T., Bania, T.M., \& Wilson, T.L. 1984, \apj, 280, 629
1682: 
1683: %\bibitem[Roshi et al. (2005)] {roshi05}
1684: %Roshi, D.A., Balser, D.S., Bania, T.M., Goss, W.M., \& De Pree, C.G. 2005, \apj, 625, 181 
1685: 
1686: %%\bibitem[Russell (1980)] {russell80} 
1687: %%Russell, R.W., Melnick, G., Gull, G.E., \& Harwit,M. 1980, \apjl, 240, L99
1688: 
1689: \bibitem[Russeil \& Castets (2004)]{russeil04}
1690: Russeil, D. \& Castets, A. 2004, \aap, 417, 107
1691: 
1692: \bibitem[Sanders et al. (1986)] {sanders86}
1693: Sanders, D.B., Clemens, D.P., Scoville, N.Z., \& Solomon, P.M.
1694: 1986, ApJS, 60, 1
1695: 
1696: \bibitem[Scoville \& Solomon (1975)] {ss75}
1697: Scoville, N. Z. \& Solomon, P. M. 1975, \apjl, 199, L105
1698: 
1699: %%\bibitem[Shah (1995)] {shah95} 
1700: %%Shah, R.Y. 1995, Senior Thesis, Boston University: ``Pressure Constraints 
1701: %%on Photodissociation Region Models via Carbon Radio Recombination Lines''
1702: 
1703: %%\bibitem[Shaver et al. (1983)] {shaver83} 
1704: %%Shaver, P.A., McGee, R.X., Newton, L.M., Danks, A.C., \& Pottasch,  
1705: %%S.R. 1983, \mnras, 204, 53
1706: 
1707: \bibitem[Simon et al. (2001)]{simon01}
1708: Simon, R., Jackson, J.M., Clemens, D.P., \& Bania, T.M. 2001, \apj, 551, 747
1709: 
1710: %%\bibitem[Simon et al. (2005), submitted] {simon05}
1711: %%Simon, R., Rathborne, J.M., Shah, R.Y., Jackson, J.M., \& Chambers,
1712: %%E.T. 2005, submitted
1713: 
1714: \bibitem[Sewilo, et al. (2004)]{sewilo04}
1715: Sewilo, M., Churchwell, E., Kurtz, S., Goss, W.M., Hofner, P. 2004, \apj, 605, 285
1716: 
1717: %%\bibitem[Sorochenko \& Tsivilev (2000)] {st00} 
1718: %%Sorochenko, R.L. \& Tsivilev, A.P. 2000, Astron. Rep., 44, 7, 426
1719: 
1720: %%\bibitem[Stacey et al. (1993)] {stacey93} 
1721: %%Stacey, G.J., Jaffe, D.T., Geis, N., Genzel, R., Harris, A.I., 
1722: %%Poglitsch, A., Stutzki, J., \& Townes, C.H. 1993, \apj, 404, 219
1723: 
1724: \bibitem[Stephenson (1992)]{stephenson92}
1725: Stephenson, C.B. 1992, \aj, 103, 263
1726: 
1727: %%\bibitem[] {} 
1728: %%St\"orzer H., Stutzki J. \& Sternberg A. 1997, \aap, 323, L13
1729: 
1730: \bibitem[Stil et al. (2006)]{stil06}
1731: Stil, J. M., et al. 2006, \aj, 132, 1158
1732: 
1733: %%\bibitem[Sullivan \& Downes (1973)] {sd73} 
1734: %%Sullivan, W.T. \& Downes, D. 1973, \aap, 29, 369
1735: 
1736: %\bibitem[Tieftrunk, Gaume, \& Wilson (1998)] {tieftrunk98}
1737: %Tieftrunk, A.R., Gaume, R.A, \& Wilson, T.L. 1998, \aap, 340, 232
1738: 
1739: %%\bibitem[Viala et al. (1988)] {viala88} 
1740: %%Viala, Y.P., Letzelter, C., Eidelsberg, M., \& Rostas, F. 1988, \aap, 
1741: %%193, 265
1742: 
1743: \bibitem[Watson et al. (2003)] {watson03}
1744: Watson, C., Araya, E., Sewilo, M., Churchwell, E., Hofner, P., \& Kurtz,
1745: 				  S. 2003, \apj, 587, 714
1746: 
1747: \bibitem[Whiteoak, Otrupcek, \& Rennie (1982)]{whiteoak82}
1748:  Whiteoak, J. B., Otrupcek, R. E., \& Rennie, C. J. 1982, PASAu, 4, 434
1749: 
1750: \bibitem[Williams, de Geus, \& Blitz (1994)]{williams94}
1751: Williams, J.P., de Geus, E. J., \& Blitz, L. 1994, \apj, 428, 693
1752: 
1753: %%\bibitem[Wilson, Tomasson \& Gardner (1975)] {wilson75} 
1754: %%Wilson, T.L., Tomasson, P., \& Gardner, F.F. 1975, \aap, 73, 167
1755: 
1756: %%\bibitem[Wolfire et al. (2003)] {wolfire03} 
1757: %%Wolfire, M.G., McKee, C.F., Hollenbach, D., \& Tielens, A.G.G.M. 2003, 
1758: %%\apj, 587, 278
1759: 
1760: \bibitem[Wood \& Churchwell (1989a)] {wc89a}
1761: Wood, D.O.S. \& Churchwell, E. 1989, \apj, 69, 831
1762: 
1763: \bibitem[Wood \& Churchwell (1989b)] {wc89b}
1764: Wood, D.O.S. \& Churchwell, E. 1989, \apj, 340, 265
1765: 
1766: \end{thebibliography}
1767: 
1768: \clearpage
1769: 
1770: \begin{deluxetable}{lcc}
1771: \tabletypesize{\scriptsize}
1772: %%\rotate
1773: \tablecaption{Faux and Anomalous Sources}
1774: \tablewidth{0pt}
1775: \tablehead{
1776: \colhead{Source} & 
1777: \colhead{Notes} & 
1778: \colhead{Reference}
1779: }
1780: 
1781: \startdata
1782: 
1783: %\input anomolous_sources.csv
1784: C14.32$+$0.13 & SNR & a \\
1785: D15.45$+$0.19 & SNR & a \\
1786: D15.52$-$0.14 & SNR & b \\
1787: U16.58$-$0.05 & No continuum peak &  \\
1788: D17.23$+$0.39 & Probably an evolved star &  c, d \\
1789: U17.64$+$0.15 & No continuum peak &  \\
1790: C18.64$-$0.29 & SNR & a, e \\
1791: U19.12$-$0.34 & No continuum peak &  \\ 
1792: U19.36$-$0.02 & No continuum peak &  \\
1793: C19.88$-$0.53 & No continuum peak &  \\
1794: C20.26$-$0.89 & No IR &  \\
1795: C20.48$+$0.17 & SNR & a, e \\
1796: D21.56$-$0.11 & SNR & a, e  \\
1797: D22.04$+$0.05 & No continuum peak &  \\
1798: D2216$-$0.16 & Star & c \\
1799: D22.40$-$0.37 & SNR & e \\
1800: C22.94$-$0.07 & No IR &  \\ 
1801: C23.07$-$0.37 & No IR  - Part of SNR? &  \\
1802: C23.07$-$0.25 & No IR  - Part of SNR? &  \\
1803: D23.16$+$0.02 & No continuum peak &  \\
1804: U23.24$-$0.24 & No continuum peak &  \\
1805: D26.47$+$0.02 & WR or LBV & f \\
1806: U26.51$+$0.28 & No continuum peak &  \\
1807: C27.13$-$0.00 & SNR & e \\
1808: C29.09$-$0.71 & SNR & e \\
1809: D29.55$+$0.11 & SNR & g \\
1810: U30.42$+$0.46 & No continuum peak &  \\
1811: D30.69$-$0.63 & No IR &  \\
1812: U30.82$+$0.27 & No continuum peak &  \\ 
1813: C30.85$+$0.13 & SNR & e \\
1814: C31.05$+$0.48 & SNR & e \\
1815: D31.61$+$0.33 & SNR & e \\
1816: D31.82$-$0.12 & SNR & e \\
1817: U33.24$+$0.01 & No continuum peak &  \\
1818: C45.48$+$0.18 & No continuum peak &  \\ 
1819: U49.67$-$0.45 & No continuum peak &  \\
1820: C50.23$+$0.33 & No IR &  \\
1821: U53.63$+$0.02 & No continuum peak &  \\
1822: 
1823: \enddata
1824: \tablenotetext{a}{\citet{brogan06}}
1825: \tablenotetext{b}{\citet{brogan06} (low confidence detection)}
1826: \tablenotetext{c}{\citet{stephenson92}}
1827: \tablenotetext{d}{\citet{kwok97}}
1828: \tablenotetext{e}{\citet{helfand06}}
1829: \tablenotetext{f}{\citet{clark03}}
1830: \tablenotetext{g}{\citet{gaensler99}}
1831: \label{tab:anomalous}
1832: \end{deluxetable}
1833: 
1834: 
1835: \begin{deluxetable}{lcccccll}
1836: \tabletypesize{\scriptsize}
1837: %%\rotate
1838: \tablecaption{\hii\ Region Source Sample}
1839: \tablewidth{0pt}
1840: \tablehead{
1841: \colhead{Source} & 
1842: \colhead{l} & 
1843: \colhead{b} & 
1844: \colhead{RA(2000.0)} & 
1845: \colhead{DEC(2000.0)} & 
1846: \colhead{$V_{\rm LSR}$} &
1847: \colhead{Reference} &
1848: \colhead{Comments} \\ 
1849: %\colhead{$\sigma_{V}$} &
1850: %\colhead{$S$} & 
1851: %\colhead{$\sigma _{S}$} & 
1852: %\colhead{$T$} & 
1853: %\colhead{$\sigma _{T}$} \\
1854: 
1855: \colhead{} & 
1856: \colhead{($\degr$)} & 
1857: \colhead{($\degr$)} & 
1858: \colhead{(h m s)} & 
1859: \colhead{($\degr$ $\arcmin$ $\arcsec$)} & 
1860: \colhead{(\kms)}  
1861: %\colhead{(\kms)} &
1862: %\colhead{(\jy)} &
1863: %\colhead{(\jy)} & 
1864: %\colhead{(mK)} & 
1865: %\colhead{(mK)}
1866: }
1867: \startdata
1868: D15.00$+$0.05a & 15.00 & $+$0.05 & 18 17 41 & $-$15 52 50 &  \phn \phn 26.5 $\pm$ 1.6 &  L96 &   \\
1869: D15.00$+$0.05b & \nodata & \nodata &  \nodata  &  \nodata  &  \phn \phn 63.5 $\pm$ 1.7 &  L96 &   \\
1870: D15.64$-$0.24 & 15.64 & $-$0.24 & 18 19 60 & $-$15 27 10 &  \phn \phn 61.8 $\pm$ 1.3 &  L96 &   \\
1871: C16.31$-$0.16 & 16.31 & $-$0.16 & 18 21 01 & $-$14 49 30 &  \phn \phn 49.5 $\pm$ 0.7 &  L89 &   \\
1872: C16.43$-$0.20 & 16.43 & $-$0.20 & 18 21 24 & $-$14 44 10 &  \phn \phn 44.5 $\pm$ 0.9 &  L89 &   \\
1873: D16.61$-$0.32 & 16.61 & $-$0.32 & 18 22 11 & $-$14 38 10 &  \phn \phn 44.9 $\pm$ 0.6 &  L89 &  a \\
1874: D16.89$+$0.13 & 16.89 & $+$0.13 & 18 21 05 & $-$14 10 30 &  \phn \phn 42.3 $\pm$ 1.6 &  L96 &   \\
1875: D17.25$-$0.20a & 17.25 & $-$0.20 & 18 22 59 & $-$14 00 50 &  \phn \phn 49.9 $\pm$ 1.4 &  L96 &   \\
1876: D17.25$-$0.20b & \nodata & \nodata &  \nodata  &  \nodata  &  \phn \phn 96.5 $\pm$ 1.9 &  L96 &   \\
1877: U18.15$-$0.28 & 18.15 & $-$0.28 & 18 25 01 & $-$13 15 20 &  \phn \phn 53.9 $\pm$ 0.4 &  L89 &   \\
1878: 
1879: \enddata
1880: 
1881: \tablenotetext{a}{Changed nebular classification from compact}
1882: 
1883: \tablecomments{Table \ref{tab:diffuse} is published in its entirety in the electronic edition of the 
1884: Astrophysical Journal Supplement Series.  A portion is shown here for
1885: guidance regarding its form and content.}
1886: 
1887: %\tablenotetext{a}{Changed nebular classification from compact}
1888: %\tablenotetext{b}{Changed from 19.67-0.14}
1889: %\tablenotetext{c}{Changed from 21.44-0.56}
1890: %\tablenotetext{d}{Changed from 23.25-0.27}
1891: %\tablenotetext{e}{Changed nebular classification from diffuse}
1892: %\tablenotetext{f}{Changed from 24.50-0.06}
1893: %\tablenotetext{g}{Changed from 25.80-0.14}
1894: %\tablenotetext{h}{Changed from 25.95+0.13}
1895: %\tablenotetext{i}{Changed from 30.06-0.32}
1896: %\tablenotetext{j}{Changed from 30.23-0.15}
1897: %\tablenotetext{k}{Changed from 30.51-0.45}
1898: %\tablenotetext{l}{Changed from 52.24-0.54}
1899: 
1900: %%tablecomment{}
1901: \label{tab:diffuse}
1902: \end{deluxetable}
1903:  
1904: \begin{deluxetable}{lcccccccccccc}
1905: \tabletypesize{\scriptsize}
1906: \rotate
1907: \tablecaption{Properties of Molecular Cloud/\hii Region Sources}
1908: \tablewidth{0pt}
1909: \tablehead{
1910: \colhead{} & 
1911: \multicolumn{5}{c}{Fitted Ellipse Parameters} & 
1912: \colhead{} &
1913: \multicolumn{3}{c}{Fitted Gaussian Parameters} &
1914: \colhead{} &
1915: \colhead{} &
1916: \colhead{} \\ \cline{2-6} \cline{8-10}
1917: 
1918: \colhead{Source} & 
1919: \colhead{l} & 
1920: \colhead{b} &
1921: \colhead{Size} &
1922: \colhead{Maj. $\times$ Min.} &
1923: \colhead{PA} &
1924: \colhead{} &
1925: \colhead{V} & 
1926: \colhead{$T_{\rm MB}$} & 
1927: \colhead{$\Delta V$} &
1928: \colhead{\tex} &
1929: \colhead{$N(^{13}{\rm CO})$} &
1930: \colhead{$CP$} \\
1931: 
1932: \colhead{} & 
1933: \colhead{($\degr$)} & 
1934: \colhead{($\degr$)} & 
1935: \colhead{($\arcmin$)} & 
1936: \colhead{($\arcmin \times \arcmin$)} & 
1937: \colhead{($\degr$)} & 
1938: \colhead{} &
1939: \colhead{(\kms)} & 
1940: \colhead{(K)} & 
1941: \colhead{(\kms)} & 
1942: \colhead{K} &
1943: \colhead{($\times 10^{16} \rm \, cm^{-2}$)} &
1944: \colhead{}
1945: }
1946: \startdata
1947: 
1948: D15.00$+$0.05a & 14.93 & $+$0.02 & \phn 2.1 & $\phn 1.6 \times  0.7\phn$ & +88.6 &   & \phn 25.85 $\pm$ 0.02 & \phn 6.03 $\pm$ 0.02 & \phn 3.77 $\pm$ 0.05 & 12.7 & \phn   3.2 & B \\
1949: D15.00$+$0.05b & \nodata & \nodata & \nodata & \nodata & \nodata &   & \phn \nodata & \phn \nodata & \phn \nodata & \nodata & \nodata & E \\
1950: D15.64$-$0.24 & 15.66 & $-$0.21 & \phn 2.0 & $\phn 2.1 \times  0.5\phn$ & +70.6 &   & \phn 56.96 $\pm$ 0.06 & \phn 1.88 $\pm$ 0.06 & \phn 1.86 $\pm$ 0.09 & \phn 8.6 & \phn   0.4 & B \\
1951: C16.31$-$0.16 & 16.36 & $-$0.21 & \phn 1.7 & $\phn 0.9 \times  0.8\phn$ & +45.0 &   & \phn 47.56 $\pm$ 0.11 & \phn 4.17 $\pm$ 0.11 & \phn 4.68 $\pm$ 0.12 & 11.0 & \phn   2.6 & B \\
1952: C16.43$-$0.20 & 16.36 & $-$0.21 & \phn 1.7 & $\phn 0.9 \times  0.8\phn$ & +14.9 &   & \phn 48.83 $\pm$ 0.02 & \phn 9.47 $\pm$ 0.02 & \phn 2.74 $\pm$ 0.04 & 13.9 & \phn   3.9 & B \\
1953: D16.61$-$0.32 & 16.56 & $-$0.34 & \phn 4.8 & $\phn 2.9 \times  2.0\phn$ & +26.3 &   & \phn 43.23 $\pm$ 0.01 & \phn 5.39 $\pm$ 0.01 & \phn 4.44 $\pm$ 0.04 & 11.1 & \phn   3.2 & C \\
1954: D16.89$+$0.13 & \nodata & \nodata & \nodata & \nodata & \nodata &   & \phn \nodata & \phn \nodata & \phn \nodata & \nodata & \nodata & D \\
1955: D17.25$-$0.20a & 17.23 & $-$0.24 & \phn 3.9 & $\phn 2.7 \times  1.4\phn$ & $-$63.4 &   & \phn 44.57 $\pm$ 0.03 & \phn 5.62 $\pm$ 0.03 & \phn 4.05 $\pm$ 0.07 & 15.7 & \phn   3.6 & B \\
1956: D17.25$-$0.20b & \nodata & \nodata & \nodata & \nodata & \nodata &   & \phn \nodata & \phn \nodata & \phn \nodata & \nodata & \nodata & D \\
1957: U18.15$-$0.28 & 18.15 & $-$0.31 & \phn 2.2 & $\phn 1.7 \times  0.7\phn$ & +77.9 &   & \phn 52.53 $\pm$ 0.04 & \phn 7.32 $\pm$ 0.04 & \phn 4.75 $\pm$ 0.06 & 24.6 & \phn   7.6 & A \\
1958: 
1959: \enddata
1960: 
1961: \tablecomments{Table \ref{tab:properties} is published in its entirety in the electronic edition of the 
1962: Astrophysical Journal Supplement Series.  A portion is shown here for
1963: guidance regarding its form and content.}
1964: 
1965: \label{tab:properties}
1966: \end{deluxetable}
1967: 
1968: \begin{deluxetable}{lccccccc}
1969: \tabletypesize{\scriptsize}
1970: %%\rotate
1971: \tablecaption{Mean Properties of Nebulae with Associated $^{13}{\rm CO}$}
1972: \tablewidth{0pt}
1973: \tablehead{
1974: \colhead{} &
1975: \colhead{N} &
1976: \colhead{$|V|$ Offset} &
1977: \colhead{$T_{\rm MB}$} &
1978: \colhead{$\Delta V$} & 
1979: \colhead{Size} &
1980: \colhead{$T_{\rm ex}$} &
1981: \colhead{$N({\rm ^{13} CO})$} \\
1982: \colhead{} &
1983: \colhead{} &
1984: \colhead{(\kms)} & 
1985: \colhead{(K)} &
1986: \colhead{(\kms)} & 
1987: \colhead{(\arcmin)} &
1988: \colhead{(K)} &
1989: \colhead{($\times 10^{16} {\,\rm cm}^{-2}$)} \\
1990: }
1991: \startdata
1992: 
1993: %\input table3.txt
1994: All & 253 &   2.98 $\pm$   2.41 &  4.77 $\pm$ 2.32 &   4.19 $\pm$  1.42 &   1.9
1995:  $\pm$  1.3 &  12.1 $\pm$  4.8 &   3.1 $\pm$  2.6 \\
1996: UC & 111 &   3.03 $\pm$   2.61 &  5.21 $\pm$ 2.40 &   4.43 $\pm$  1.38 &   1.7
1997:  $\pm$  1.1 &  12.2 $\pm$  5.0 &   3.5 $\pm$  2.8 \\
1998: Compact & \phn95 &   3.01 $\pm$   2.37 &  4.96 $\pm$ 2.36 &   4.23 $\pm$  1.40
1999:  &   2.2 $\pm$  1.6 &  13.3 $\pm$  4.7 &   3.3 $\pm$  2.6 \\
2000: Diffuse & \phn47 &   2.81 $\pm$   2.01 &  3.32 $\pm$ 1.29 &   3.56 $\pm$  1.37
2001:  &   1.9 $\pm$  1.0 & \phn  9.3 $\pm$  3.0 &   1.6 $\pm$  1.0 \\
2002: A & 112 &   2.65 $\pm$   2.06 &  5.63 $\pm$ 2.51 &   4.52 $\pm$  1.26 &   1.7
2003:  $\pm$  0.8 &  12.8 $\pm$  5.1 &   4.0 $\pm$  3.0 \\
2004: B & \phn 75 &   3.00 $\pm$   2.37 &  4.62 $\pm$ 2.06 &   4.05 $\pm$  1.31 & 
2005:   2.1 $\pm$  1.5 &  12.1 $\pm$  4.4 &   2.8 $\pm$  2.0 \\
2006: C & \phn 66 &   3.53 $\pm$   2.90 &  3.46 $\pm$ 1.51 &   3.80 $\pm$  1.66 & 
2007:   2.2 $\pm$  1.7 &  11.0 $\pm$  4.4 &   1.9 $\pm$  1.9 \\
2008: 
2009: \enddata
2010: 
2011: %\centerline{\bf Errors are standard deviation: $1 \sigma $}
2012: 
2013: \label{tab:groups}
2014: \end{deluxetable}
2015: 
2016: 
2017: \clearpage
2018: 
2019: 
2020: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2021: \begin{figure}
2022: \epsscale{1.0}
2023: \plotone{f1.eps}
2024: 
2025: \caption{Longitude-LSR velocity diagram for \hii\ regions located inside
2026: the GRS survey zone. The nebulae are shown projected onto the Galactic
2027: plane.  The symbols represent UC nebulae (small filled circles), compact
2028: nebulae (medium filled circles), and diffuse nebulae (large open circles).}
2029: 
2030: \label{fig:lv}
2031: \end{figure}
2032: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2033: 
2034: \begin{figure}
2035: \epsscale{1.0} 
2036: \plotone{f2.eps}
2037: \caption{The CO/\hii\ association procedure for the U43.24$-$0.05 
2038: \hii\ region (see text). 
2039: %
2040: Panel A: Using single velocity channels near the RRL velocity and
2041: nominal \hii\ region position we first locate the associated molecular
2042: emission.  The black cross shows the \hii\ region position.  The black
2043: rectangle defines the extent of the \cor\ emission that is deemed to be
2044: associated with the \hii\ region.
2045: %
2046: Panel B: The average \cor\ spectrum of the GRS data cube voxels that lie
2047: within the panel A black rectange is shown.  The solid vertical line flags the
2048: \hii\ region RRL velocity and the dashed lines show the velocity range of the associated emission.
2049: %
2050: Panel C: The integrated intensity image made from the data cube using
2051: the line center velocity and FWHM line width found by the Panel B
2052: Gaussian fit is shown.  This image is then used to define the final
2053: extent of \cor\ emission that is deemed to be associated with the \hii\
2054: region.  The black outline shows this region which is defined by a
2055: threshold algorithm (see text).
2056: %
2057: Panel D: The average \cor\ spectrum of the data cube voxels that lie
2058: within the panel C threshold defined region is shown. The vertical line
2059: flags the RRL velocity.  The  Gaussian fit to the emission
2060: line is used to derive the physical properties of the \cor\ gas
2061: associated with this \hii\ region.  The spectrum from panel B is shown in gray for comparison.
2062: }
2063: 
2064: \label{fig:3plot}
2065: \end{figure}
2066: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2067: 
2068: \begin{figure}
2069: \epsscale{1.0} 
2070: \plotone{f3.eps} 
2071: 
2072: \caption{Integrated intensity images (grey scale and black contours) 
2073: for a sample of \hii\ regions with A confidence parameters. Nominal
2074: \hii\ region positions are marked with crosses and our fitted ellipses
2075: are also shown.  Contours of 21cm VGPS continuum emission are
2076: shown in grey.  Tickmarks on the VGPS contours point downhill, towards decreasing values.  See
2077: \S \ref{sec:discussion} for a detailed description of these images.}
2078: 
2079: \label{fig:images_a}
2080: \end{figure}
2081: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2082: 
2083: \begin{figure}
2084: \epsscale{1.0} 
2085: \plotone{f4.eps}  
2086: 
2087: \caption{Integrated intensity images for a sample of \hii\ regions with
2088: B confidence parameters.  Other parameters are as in Figure \ref{fig:images_a}.}
2089: 
2090: \label{fig:images_b}
2091: \end{figure}
2092: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2093: 
2094: \begin{figure}
2095: \epsscale{1.0} 
2096: \plotone{f5.eps}  
2097: 
2098: \caption{Integrated intensity images for a sample of \hii\ regions with
2099: C confidence parameters (top row), D (lower left) and E (lower
2100: right). Fitted ellipses are shown for C sources only.  Other
2101: parameters are as in Figure \ref{fig:images_a}.}
2102: 
2103: \label{fig:images_cde}
2104: \end{figure}
2105: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2106: 
2107: \begin{figure}
2108: \epsscale{1.0} 
2109: \plotone{f6.eps}
2110: 
2111: \caption{The number distributions of the qualitative confidence
2112: parameter, $CP$.  Sources with associated molecular gas are rated A, B
2113: or C in order of decreasing confidence in the association (see \S \ref{sec:discussion} text). Confidence
2114: parameter D and E sources have little or no associated molecular
2115: emission, respectively. The top left panel is a stacked histogram
2116: where the top line shows the histogram for the entire sample of 301
2117: \hii\ regions.  The open, hatched and grey histograms show the
2118: contribution that UC, Compact, and Diffuse nebulae, respectively, make
2119: to the total in each bin.}
2120: 
2121: \label{fig:confidence}
2122: \end{figure}
2123: 
2124: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2125: \begin{figure}
2126: \epsscale{0.90}
2127: \plotone{f7.eps}
2128: \caption{Longitude-LSR velocity diagram for 253 \hii\ regions in our
2129: sample that show associated \cor\ emission.  Symbols indicate the
2130: confidence parameter of each nebula: $CP$ A (large filled circles), B
2131: (small filled circles), or C (small open circles).  The lines show the
2132: locii of the LSR terminal velocity expected from two different
2133: Galactic rotation curve models: \citet{clemens85} (dotted line) and
2134: \citet{brand86} (full line).  }
2135: 
2136: \label{fig:co_lv}
2137: \end{figure}
2138: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2139: 
2140: \begin{figure}
2141: \epsscale{1.0}
2142: \plotone{f8.eps}
2143: 
2144: \caption{Nebular excitation temperature, \tex, derived from the \co\, survey
2145: of \citet{sanders86}. Panels are the same as in Figure
2146: \ref{fig:confidence}.}
2147: 
2148: \label{fig:t_ex}
2149: \end{figure}
2150: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2151: 
2152: \begin{figure}
2153: \epsscale{1.0}
2154: \plotone{f9.eps}
2155: 
2156: \caption{Estimated H$_2$ column density (top axis) and \cor\, column
2157: density (bottom axis) for the molecular clumps associated with our \hii\
2158: regions. Panels are the same as in Figure \ref{fig:confidence}.}
2159: 
2160: \label{fig:column}
2161: \end{figure}
2162: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2163: 
2164: \begin{figure}
2165: \epsscale{1.0}
2166: \plotone{f10.eps}
2167: 
2168: \caption{The LSR velocity difference between the \hii\ region RRL and
2169: molecular gas \cor\, velocity. The distribution peaks near zero, as
2170: expected. Panels are the same as in Figure \ref{fig:confidence}.}
2171: 
2172: \label{fig:vel_diff_by_code}
2173: \end{figure}
2174: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2175: 
2176: \begin{figure}
2177: \epsscale{1.0}
2178: \plotone{f11.eps}
2179: 
2180: \caption{Main beam brightness temperature \cor\, line intensity from Gaussian fits. 
2181: Panels are the same as in Figure \ref{fig:confidence}.}
2182: 
2183: \label{fig:intensity}
2184: \end{figure}
2185: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2186: 
2187: \begin{figure}
2188: \epsscale{1.0}
2189: \plotone{f12.eps}
2190: 
2191: \caption{\cor\, FWHM line width from Gaussian fits.  
2192: Panels are the same as in Figure \ref{fig:confidence}.}
2193: 
2194: \label{fig:linewidth}
2195: \end{figure}
2196: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2197: 
2198: \begin{figure}
2199: \epsscale{1.0}
2200: \plotone{f13.eps}
2201: 
2202: \caption{Angular size defined as the geometric mean diameter of the fitted 
2203: ellipses. Panels are the same as in Figure \ref{fig:confidence}.}
2204: 
2205: \label{fig:angsize}
2206: \end{figure}
2207: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2208: 
2209: %\begin{figure}
2210: %\epsscale{.80}
2211: %\plotone{concentration_groups_4panel.ps} 
2212: 
2213: %\caption{Concentration of our associated molecular clumps from our
2214: %fitted ellipses. Panels are the same as in Figure \ref{fig:confidence}.}
2215: 
2216: %\label{fig:concentration_diffuse}
2217: %\end{figure}
2218: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2219: 
2220: \begin{figure}
2221: \epsscale{1.0}
2222: \plotone{f14.eps}
2223: 
2224: \caption{\cor\, line intensity plotted as a function of FWHM line
2225: width.  Shown are all the $CP$ A, B, or C sources: UC (filled triangles),
2226: Compact (filled circles), and Diffuse (open squares) nebulae.  The lines
2227: dividing the figure into quadrants are from an analysis of molecular
2228: clumps in the GRS dataset (see \S \ref{sec:GRS}).}
2229: 
2230: \label{fig:t_vs_fwhm}
2231: 
2232: \end{figure}
2233: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2234: 
2235: %\begin{deluxetable}{lccccccc}
2236: %\tabletypesize{\scriptsize}
2237: %%\rotate
2238: %\tablecaption{Mean Properties of Nebulae with Associated $^{13}{\rm CO}$}
2239: %\tablewidth{0pt}
2240: %\tablehead{
2241: %\colhead{} &
2242: %\colhead{N} &
2243: %\colhead{$|V|$ Offset} &
2244: %\colhead{$T_{\rm MB}$} &
2245: %\colhead{$\Delta V$} & 
2246: %\colhead{Size} &
2247: %\colhead{$T_{\rm ex}$} &
2248: %\colhead{$N({\rm ^{13} CO})$} \\
2249: %\colhead{} &
2250: %\colhead{} &
2251: %\colhead{(\kms)} & 
2252: %\colhead{(K)} &
2253: %\colhead{(\kms)} & 
2254: %\colhead{(\arcmin)} &
2255: %\colhead{(K)} &
2256: %\colhead{($\times 10^{16} {\,\rm cm}^{-2}$)} \\
2257: %}
2258: %\startdata
2259: 
2260: %\input table3_sem.txt
2261: 
2262: %\enddata
2263: 
2264: %\centerline{\bf Errors are standard error of the mean: $\sigma/\sqrt{N}$}
2265: 
2266: %\label{tab:sem}
2267: %\end{deluxetable}
2268: 
2269: \end{document}
2270: 
2271: 
2272: