Where is OH and Does It Trace the Dark Molecular Gas (DMG)?
Di Li, Ningyu Tang, Hiep Nguyen, J. R. Dawson, Carl Heiles, Duo Xu, Zhichen Pan, Paul F. Goldsmith, Steven J. Gibson, Claire E. Murray, Tim Robishaw, N. M. McClure-Griffiths, John Dickey, Jorge Pineda, Snežana Stanimirović, L. Bronfman, Thomas Troland, PRIMO collaboration
aa r X i v : . [ a s t r o - ph . GA ] J a n Preprint typeset using L A TEX style emulateapj v. 12/16/11
WHERE IS OH AND DOES IT TRACE THE DARK MOLECULAR GAS (DMG)?
Di Li , Ningyu Tang , Hiep Nguyen , J. R. Dawson , Carl Heiles , Duo Xu , Zhichen Pan , Paul F.Goldsmith , Steven J. Gibson , Claire E. Murray , Tim Robishaw , N. M. McClure-Griffiths , JohnDickey , Jorge Pineda , Sneˇzana Stanimirovi´c , L. Bronfman , Thomas Troland , and the PRIMOcollaboration ABSTRACTHydroxyl (OH) is expected to be abundant in diffuse interstellar molecular gas as it forms alongwith H under similar conditions and within a similar extinction range. We have analyzed absorp-tion measurements of OH at 1665 MHz and 1667 MHz toward 44 extragalactic continuum sources,together with the J=1-0 transitions of CO, CO, and C O, and the J=2-1 of CO. The excitationtemperature of OH were found to follow a modified log-normal distribution f ( T ex ) ∝ √ πσ exp (cid:20) − [ln( T ex ) − ln(3 . σ (cid:21) , the peak of which is close to the temperature of the Galactic emission background (CMB+synchron).In fact, 90% of the OH has excitation temperature within 2 K of the Galactic background at the samelocation, providing a plausible explanation for the apparent difficulty to map this abundant moleculein emission. The opacities of OH were found to be small and peak around 0.01. For gas at intermediateextinctions (A V ∼ N (OH) ≃ cm − is approximately independent of A V . We conclude that OH is abundant in the diffuse molecular gasand OH absorption is a good tracer of ‘dark molecular gas (DMG)’. The measured fraction of DMGdepends on assumed detection threshold of the CO data set. The next generation of highly sensitivelow frequency radio telescopes, FAST and SKA, will make feasible the systematic inventory of diffusemolecular gas, through decomposing in velocity the molecular (e.g. OH and CH) absorption profilestoward background continuum sources with numbers exceeding what is currently available by ordersof magnitude. Subject headings:
ISM: clouds — ISM: evolution — ISM: molecules. National Astronomical Observatories, CAS, Beijing 100012,China; Email: [email protected], [email protected] Key Laboratory of Radio Astronomy, Nanjing, ChineseAcademy of Science University of Chinese Academy of Sciences, Beijing 100049,China Department of Physics and Astronomy and MQ ResearchCentre in Astronomy, Astrophysics and Astrophotonics, Mac-quarie University, NSW 2109, Australia Australia Telescope National Facility, CSIRO Astronomyand Space Science, PO Box 76, Epping, NSW 1710, Australia Department of Astronomy, University of California, Berke-ley, 601 Campbell Hall 3411, Berkeley, CA 94720-3411 Department of Astronomy, The University of Texas atAustin, Austin, TX 78712, USA Jet Propulsion Laboratory, California Institute of Technol-ogy, 4800 Oak Grove Drive, Pasadena, CA 91109, USA Western Kentucky University, Dept. of Physics and Astron-omy, 1906 College Heights Blvd, Bowling Green, KY 42101,USA University of Wisconsin, Department of Astronomy, 475 NCharter St., Madison, WI 53706, USA Space Telescope Science Institute, 3700 San Martin Drive,Baltimore, MD 21218, USA Dominion Radio Astrophysical Observatory, National Re-search Council, PO Box 248, Penticton,BC, V2A 6J9, Canada Research School for Astronomy & Astrophysics, AustralianNational University, Canberra, ACT 2611, Australia University of Tasmania, School of Maths and Physics,Hobart, TAS 7001, Australia Departamento de Astronom´ıa, Universidad de Chile,Casilla 36, Santiago de Chile, Chile Department of Physics and Astronomy, University ofKentucky, Lexington, Kentucky 40506 Pacific Rim Interstellar Matter Observers; http://ism.bao.ac.cn/primo
Li et al. INTRODUCTION
The two relatively denser phases of the interstellarmedium (ISM) are the atomic Cold Neutral Medium(CNM) traced by the H i λ ) clouds, usuallytraced by CO. CO has historically been the most impor-tant tracer of molecular hydrogen, which remains largelyinvisible due to its lack of emission at temperatures inthe molecular ISM. Empirically, CO intensities have beenused as an indicator of the total molecular mass in theMilky Way and external galaxies through the so-called“X-factor”, with numerous caveats, not least of which isthe large opacities of CO transitions. Gases in these twophases dominate the masses of star forming clouds on agalactic scale. The measured ISM gas mass from H i andCO is thus the foundation of many key quantities in un-derstanding galaxy evolution and star formation, such asthe star formation efficiency.A growing body of evidence, however, indicates the ex-istence of gas traced by neither H i nor CO. Comparativestudies (e.g., de Vries et al. 1987) of Infrared AstronomySatellite (IRAS) dust images and gas maps in H i andCO revealed an apparent ‘excess’ of dust emission. ThePlanck Collaboration et al. (2011) clearly showed excessdust opacity in the intermediate extinction range A V ∼ and CO, respectively. The missinggas, or rather, the undetected gas component, is widelyreferred to as dark gas, popularized as a common termby Grenier et al. (2005). These authors found more dif-fuse gamma-ray emission observed by Energetic GammaRay Experiment Telescope (EGRET) than can be ex-plained by cosmic-ray interactions with the observed H-nuclei. Observations of the THz fine structure C + linealso helped reveal the dark gas, from which the C + linestrength is stronger than can be produced by only H i gas (Langer et al. 2010; Pineda et al. 2013; Langer et al.2014). A minority of the ISM community have arguedthat dark gas can be explained by underestimated H i opacities (Fukui et al. 2015), which is in contrast withsome other recent works (Stanimirovi´c et al. 2014; Leeet al. 2015). We focus here on the dark molecular gas(DMG), or more specifically CO-dark molecular gas.ISM chemistry and PDR models predict the existenceof H in regions where CO is not detectable (Wolfire etal. 2010). CO can be of low abundance due to photo-dissociation in unshielded regions and/or can be heavilysub-thermal due to low collisional excitation rate in thediffuse gas. OH, or Hydroxyl, was the first interstellarmolecule detected at radio wavelenghs (Weinreb et al.1963). It can form efficiently through relatively rapidroutes including charge-exchange reactions initiated bycosmic ray ionization once H becomes available (vanDishoeck & Black 1988). Starting from H + ,O + H + → O + + H , O + + H → OH + + H . (1)Also starting from H +2 ,H +2 + H → H +3 + H , H +3 + O → OH + + H . (2)OH + then reacts with H to form H O + that continues on to H O + , OH + + H → H O + + H , H O + + H → H O + + H . (3)H O + and H O + recombine with electrons to form OH.OH can join the carbon reaction chain through reactionwith C + and eventually produce CO,OH + C + → CO + + H , CO + + H → HCO + + H , HCO + + e − → CO + H . (4)Widespread and abundant OH, along with HCO + andC + is thus expected in diffuse and intermediate extinc-tion regions.OH has been widely detected throughout the Galacticplane (e.g., Caswell & Haynes 1975; Turner 1979; Boyce& Cohen 1994; Dawson et al. 2014; Bihr et al. 2015), inlocal molecular clouds (e.g., Sancisi et al. 1974; Wouter-loot & Habing 1985; Harju et al. 2000), and in high-latitude translucent and cirrus clouds (e.g., Grossmannet al. 1990; Barriault et al. 2010; Cotten et al. 2012). Cru-cially, a small number of studies have confirmed OH ex-tending outside CO-bright regions (Wannier et al. 1993;Liszt & Lucas 1996; Allen et al. 2012, 2015), and/or as-sociated with narrow H i absorption features (Dickey etal. 1981; Liszt & Lucas 1996; Li & Goldsmith 2003), con-firming its viability as a dark gas tracer. OH and HCO + have been shown to be tightly correlated in absorptionmeasurements against extragalactic continuum sources(Liszt & Lucas 1996; Lucas & Liszt 1996).Because OH in emission is typically very weak, large-scale OH maps remain rare, particularly for diffusemolecular gas, which should presumably be dominatedby DMG. Detectability often hinges on the presence ofbright continuum background against which the OH linescan be seen in absorption, either the bright diffuse emis-sion of the inner Galactic plane (e.g., Dawson et al. 2014)or bright, compact extragalactic sources (e.g., Goss 1968;Nguyen-Q-Rieu et al. 1976; Crutcher 1977, 1979; Dickeyet al. 1981; Colgan et al. 1989; Liszt & Lucas 1996). Thislatter approach has the additional advantage that on-source and off-source comparison can be made directly toderive optical depths and excitation temperatures. Thisis the approach we take in this work.Heiles & Troland (2003a,b) published the MillenniumSurvey of 21-cm line absorption toward 79 continuumsources. The ON-OFF technique and Gaussian decom-position analysis allowed them to provide direct measure-ments of the excitation temperature and column densityof H i components. Given that the Millennium sources aregenerally out of the Plane, the absorption componentsare biased toward local gas. The large gain of Areciboand the substantial integration time spent on each sourcemade the Millennium Survey one of the most sensitivesurveys of the diffuse ISM. Among the significant find-ings is the fact that a substantial fraction of the CNMlies below the canonical 100K temperature predicted byphased ISM models (Field et al. 1969; McKee & Ostriker1977) for maintaining pressure balance. The existence ofcold gas in significant quantities points to the necessityof utilizing absorption measurements for a comprehensivecensus of ISM, taking into account the general Galactichere is OH and Does It Trace the Dark Molecular Gas (DMG)? 3radiation field.The L-wide receiver at Arecibo allows for simultane-ous observation of H i and OH. This was carried out bythe Millennium Survey, but the OH data have remainedunpublished until now. To analyze these OH absorp-tion measurements in the context of DMG, we conducted3mm and 1mm CO observations toward the Millenniumsources and performed a combined analysis of their exci-tation and abundances.This paper is organized as follows: In section 2, wedescribe the observations of H i , OH, and CO. In section3, we analyze the OH line excitations and other proper-ties. In section 4, we explore the relation between thesethree spectral tracers. Discussions and conclusions arepresented in section 5 and section 6, respectively. OBSERVATIONS H i and OH During the Millennium Survey, the Λ-doubling transi-tions of ground-state OH at 1665.402 and 1667.359 MHzwere obtained simultaneously with H i with the AreciboL-wide receiver towards 72 of the 79 survey positions.These sources typically had flux density S . & − . An RMS of28 mK (T A ) per channel was achieved with 2 hours oftotal integration time. Twenty-one sightlines exhibitedOH absorption. CO We conducted a follow-up CO survey of 44 of the Mil-lennium sight-lines for which OH data were taken. Fig. 1shows the distribution of all observed sources in Galacticcoordinates.The J=1–0 transitions of CO, CO, and 1¸8o were ob-served with the Purple Mountain Observatory Delingha(PMODLH) 13.7 m telescope of the Chinese Academyof Sciences. All numbers reported in this section arein the units of T mb , since the Delingha system automati-cally corrected for the main beam antenna efficiency. Thethree transitions were observed simultaneously with the3 mm SIS receiver in March 2013, May 2013, May 2014,and May 2016. The FFTS wide-band spectral backendhas a bandwidth of 1 GHz at a frequency resolution of61.0 kHz, which corresponds to 0.159 km s − at 115.0GHz and 0.166 km s − at 110.0 GHz. Position-switchingmode was used with reference positions selected from theIRAS Sky Survey Atlas . The system temperature var-ied from 210 K to 350 K for CO, and 140 K to 225 Kfor the CO and C O observations. The resulting RMS http : //irsa.ipac.caltech.edu/data/ISSA/ Fig. 1.—
The location of background continuum sources in Galac-tic coordinates. This plot shows only the 44 sources towards whichH i OH and CO were observed. Open circles represent sourceswith detected H i absorption only. Squares represent sources withdetected H i and OH absorption. Red triangles represent sourceswith CO detections, in which every detected OH component isalso seen in CO. Blue dots represent sources with CO detections,in which some OH components do not have corresponding CO de-tections. We call these sources “partial CO detections”. are ∼
60 mK for a 0.159 km s − channel for CO and ∼
30 mK for a 0.166 km s − channel for CO and C O,respectively.The CO(J=2–1) data were taken with the CaltechSubmillimeter Observatory (CSO) 10.4 m on Mauna Keain July, October, and December of 2013. The systemtemperature varied from 230 to 300 K for CO(J=2–1),resulting in an RMS of ∼
35 mK at a velocity resolutionof 0.16 km s − .To achieve consistent sensitivity among sight-lines, CO(J=2–1) spectra toward 3 sources were also ob-tained with the IRAM 30m telescope in frequency-switching mode on the 22nd and 23rd of May, 2016. Theintegration times of these observation were between 30and 90 minutes, resulting in an RMS of less than 20 mKat a velocity resolution of 0.25 km s − .The astronomical software package Gildas/CLASS was used for data reduction including baseline removaland Gaussian fitting. OH PROPERTIES
Radiative Transfer and Gaussian Analysis
The equations of radiative transfer for ON/OFF sourcemeasurements may be written as T ONA ( v ) /η b = ( T ex − T bg − T c )(1 − e − τ v ) (K) , (5) T OFFA ( v ) /η b = ( T ex − T bg )(1 − e − τ v ) (K) , (6)where we assumed main beam efficiency η b = 0 . T ex and τ v are excitation temperature and optical depthof the cold cloud, respectively. T ONA ( v ) and T OFFA ( v ) areantenna temperatures toward and offset from the contin-uum source, respectively. Here, T c is the compact contin-uum source brightness temperature, and T bg is the back-ground brightness temperature, consisting of the 2.7 Kisotropic CMB and the Galactic synchrotron background http : Li et al.at the source position. We adopted the same treatmentof the background continuum (Heiles & Troland 2003a) T bg = 2 . T bg408 ( ν OH /
408 MHz) − . , (7)where T bg comes from Haslam et al. (1982). Thebackground continuum contribution from Galactic H ii regions can be safely ignored, since our sources are eitherat high Galactic latitudes or Galactic Anti-Center longi-tudes. Typical T bg values are thus found to be around3.3 K.We decomposed the OH spectra into Gaussian compo-nents to evaluate the physical properties of OH cloudsalong the line-of-sight. Following the methodology ofHeiles & Troland (2003a), we assumed a two-phasemedium, in which cold gas components are seen in bothabsorption and emission (i.e. in both the opacity andbrightness temperature profiles), while warm gas appearsonly in emission, i.e. only in brightness temperature (seeHeiles & Troland (2003a) for further details). While thistechnique is generally applicable for both H i and OH,we have only detected OH in absorption in this work.“Warm” OH components in emission have been observedby Liszt & Lucas (1996), although only by the NRAO43m and not with the VLA or Nancay in their study.This is consistent with our results in that there is noOH warm enough ( > a few hundred K) in our Arecibobeam to be seen in emission, nor do we expect it fromastrochemistry considerations.In brief, the expected profile T exp ( v ) consists of bothemission and absorption components: T exp ( v ) = T B , cold ( v ) + T B , warm ( v ) , (8)where T B , cold ( v ) is the brightness temperature of the coldgas and T B , warm ( v ) is the brightness temperature of thethe warm gas. Both components contribute to the emis-sion profile.The opacity spectrum is obtained by combining the on-and off-source spectra (equations 5 and 6), and containsonly cold gas seen in absorption. First, we fit the ob-served opacity spectrum e − τ ( v ) with a set of N Gaussiancomponents.Next we fit the expected emission profile, T exp , whichis assumed to also consist of the N cold components seenin the absorption spectrum, plus any warm componentsseen only in emission. We further assume that each com-ponent is independent and isothermal with an excitationtemperature T ex,n : τ ( v ) = N − X n =0 τ ,n e − [( v − v ,n ) /δv n ] . (9)Here τ ,n , v ,n , δv n are respectively the peak opticaldepth, central V LSR , and 1/ e -width of component n .All the values of τ ,n , v ,n and δv n were then obtainedthrough least-squares fitting. The contribution of thecold components is given by T B , cold ( v ) = N − X n =0 T ex ,n (1 − e − τ n ( v ) ) e − n − P m =0 τ m ( v ) , (10)where subscript m describes the M absorbing clouds ly-ing in front of the n th cloud. When n=0, the summationover m takes no effect, as there is no foreground cloud. e − τ P0428 + − τ = [0.0015, 0.0076]VLSR = [3.60, 10.70]FWHM = [1.06, 1.10]T ex = [8.424, 8.420] − R es i du a l s − − T ex p ( K ) VLSR (km / s) − − R es i du a l s Fig. 2.—
Illustration of derived parameters from the fits to theabsorption (upper panel) and expected emission (lower panel) pro-files for the source P0428+20. The thin solid lines show the data;the thick solid lines are the fits. The solid lines below the dashed-lines present the residuals for absorption and fitted expected emis-sion, respectively.
We obtained the values of excitation temperature T ex ,n from this fit. As in Heiles & Troland (2003a), we exper-imented with all possible orders along the line of sightand retained the one that yields the smallest residuals.In total, we detected 48 OH components towards 21sightlines. Example spectra and fits are shown in Figures2 and 3. OH Excitation and Optical Depth
Our fitting scheme provides measurements of the exci-tation conditions and optical depths of the OH gas.Figure 4 shows a histogram of optical depths in thetwo OH main lines, with the 1665 line scaled by a factorof 9/5 (see below). The measured optical depth of bothlines peaks at very low values of ∼ .
01, with a longertail extending as high as 0.21 in the 1667 line. As canbe seen in Table 2, the uncertainties on these values aregenerally very small. The OH gas probed by absorptionis thus quite optically thin.As seen in Figure 5, OH excitation temperatures inthe two main lines peak at ∼ ∼ T ex values within 2 K of the diffuse continuumbackground temperature, T bg , consistent with measure-ments from past work (e.g., Nguyen-Q-Rieu et al. 1976;Crutcher 1979; Dickey et al. 1981). The probability den-sity distribution (PDF) of the OH excitation tempera-ture can be fitted by a modified normalized lognormalhere is OH and Does It Trace the Dark Molecular Gas (DMG)? 5 e − τ − τ = [0.0062, 0.0199]VLSR = [14.63, 15.40]FWHM = [1.86, 0.89]T ex = [7.96, 2.35] − − R es i du a l s − − T ex p ( K ) VLSR (km / s) − − R es i du a l s Fig. 3.—
Data for the source 3C409. See Fig.2 for completedescription.
Fig. 4.—
Histogram of peak optical depths of the OH Gaussiancomponents. Black shows the results for the 1667 MHz line andred shows 1.8 times the results for the 1665 MHz line.
Fig. 5.—
Histograms of excitation temperature, T ex , of the Gaus-sian components for the two OH main lines and the backgroundcontinuum temperature T bg . The number histogram of T bg hasbeen scaled by a factor of 0.5, for ease of visualization. function , f ( T ex ) ∝ √ πσ exp (cid:20) − [ln( T ex ) − ln( T )] σ (cid:21) . (11)The fit parameters are sensitive to the binning and sta-tistical weighting of the data and are not unique. Wechose one solution that preserves the tail of relativelyhigh T ex . The exact numerical values are less meaningfulthan the location of the distribution peak and the roughtrend of T ex , which are represented in the current fit-ting. The fit results are shown in Figure 6 and Table 1.The statistical uncertainties are small. Similar numericalvalues are found in fitting both the 1665 and 1667 transi-tions. A simple average of the two sets of fit parametersis also presented in Table 1. Given the unbiased natureof absorption-selected sightlines and the fact that almosthalf of the detected OH components lie at low GalacticLatitudes ( | b | < ◦ ), such a generic distribution functioncan may be representative of OH excitation conditionsin the Galactic ISM. This tentative conclusion will berefined by future observations.The combined results of T ex and τ reflect the complex-ities in OH excitation. When the level populations ofthe OH ground states are in LTE, the excitation tem- Equation 11 differs from standard log-normal function in thatit has no variable (x) in the denominator.
Li et al.
Fig. 6.—
Fitting results of the PDFs of OH excitation tempera-ture for both the 1665 and 1667 lines. Fit parameters are given inTable 1.
TABLE 1Lognormal fit parameters for theOH T ex distribution, as shown inFigure 6. Line Fitted T a Fitted σ a OH 1665 3.4 0.98OH 1667 3.2 0.96OH average 3.3 0.97 a Parameter defined in Equation 11. peratures of the 1665 and 1667 MHz lines are equal, andtheir optical depth ratio ( R / = τ /τ ) is 1.8. Ingeneral, however, we do not expect OH to be thermallyexcited. The satellite lines (at 1612 and 1720 MHz) com-monly show highly anomalous excitation patterns, with T ex that are strongly subthermal in one line, and eithervery high or negative (masing) in the other (e.g., Dawsonet al. 2014). This may occur due to far-IR or infraredpumping, which can lead to selective overpopulation ofeither the F = 1 or the F = 2 level pair (e.g., Guibertet al. 1978). The result is that the main line T ex may bevery similar, despite the system as a whole being stronglynon-thermally excited.As shown in Figure 7, most OH components deviatefrom LTE at more than the formal 1 σ uncertainties prop-agated from the Gaussian fits. The difference betweenthe excitation temperatures, however, mainly falls within2 K ( | ∆ T ex | < OH Column Density
The column densities of the OH components were com-puted independently from each line according to: N (OH) = 8 πk T ex , ν A c h Z τ dv , (12) N (OH) = 8 πk T ex , ν A c h Z τ dv , (13)where A = 7 . × − s − and A = 7 . × − s − are the Einstein A-coefficients of the OH main − − − − − ∆ T ex = T ex (1667) - T ex (1665) (K)1.01.52.02.53.0 R / = τ / τ Fig. 7.—
Optical depth ratio ( R / = τ /τ ) as a func-tion of excitation temperature difference ( T ex (1667) − T ex (1665))for the 1667 and 1665 MHz OH lines. The horizontal and verti-cal dashed lines indicate 1.8 and 0.0, the values for LTE excitation.Error bars indicate the 1 σ formal uncertainties propagated throughfrom the Gaussian fits. Black points are those consistent with LTEto within the 1 σ errors; red points are inconsistent. The verticalshaded region represents | ∆ T ex | < lines (Destombes et al. 1977). The values are tabulatedin Table 2 and discussed below. THE RELATION BETWEEN H I , CO, AND OH OH and H i Regardless of the complexities in the OH excitation,the measured line ratios are generally close to the LTEvalue of 5/9 with a relatively small scatter. To appropri-ately utilize both transitions while minimizing the im-pact of the poorer S/N in the 1665 line, we estimatethe total OH column density as N (OH) = N (1665) + N (1667). The OH column densities obtained with thismethod are plotted against H i column density of eachGaussian component in Figure 8 on a component-by-component basis. Since the H i components are alwayswider than OH lines, the sum of all OH components wasused when multiple OH components coincide with oneH i component.For non-detections, an upper limit was estimated basedon the 1667 MHz spectrum alone, assuming a singleGaussian optical depth spectrum, with a FWHM of 1.0km/s and peak τ equal to 3 times the spectral RMS.For instance, rms of 1667 MHz absorption spectrum to-ward 3C138 is 31 mK in brightness temperature. T ex was assumed to be equal to 3.5 K, the peak of the log-normal function fitted to the T ex distribution in Figure6. These values are plotted as triangles in Figure 8. Ourabsorption data set typically has a detection limit around N (OH) ≃ cm − , with a number of higher limits oc-curring towards weaker continuum background sources.Many H i components have no detectable OH. WhereOH is detected, there is some suggestion of a weak cor-relation between the OH and H i column densities. Forhere is OH and Does It Trace the Dark Molecular Gas (DMG)? 7 − N HI (10 cm − )10 − − N O H ( c m − ) Upper limit N OH (1667)N OH (1667) Fig. 8.—
OH column densities derived from the 1667 line, N (OH), versus H i column densities, N (HI), both on a Gaussiancomponent-by-component basis. The triangles depict approximateupper limits for H i components where OH was not detected (seeSection 4.1). N (HI) between 10 and 10 cm − most sources areconsistent with an [OH]/[HI] abundance ratio ∼ − . OH and CO
Both CO and OH are widely used molecular tracersin the ISM. Allen et al. (2012, 2015) performed a pilotOH survey toward the Galactic plane around l ≈ ◦ and demonstrated the presence of CO-dark molecular gas(DMG): CO emission was absent in more than half ofthe detected OH spectral features (see also Wannier etal. (1993) and Barriault et al. (2010)). Xu & Li (2016)took OH observations across a boundary of the Taurusmolecular cloud, revealing that the fraction of DMG de-creases from 0.8 in the outer CO poor region to 0.2 inthe inner CO abundant region.Our explicit measurements of CO and OH propertiesallow us to examine the relation between CO and OHagain. Nine OH components are identified as DMGclouds and will be discussed in detail in section 4.3. Wefocus here on molecular clouds with both OH and COdetections.CO emission was detected toward 40/49 OH compo-nents. CO column density, N (CO), is calculated differ-ently for each of the following three cases:1. Detection of only CO(J=1–0).2. Detection of both CO(J=1–0) and CO(J=2–1).3. Detection of three lines, CO(J=1–0), CO(J=2–1), and CO(J=1–0) simultaneously.In LTE, the total column density N tot of a two-level transition from upper level u to lower level l is given by N tot = 8 πν c A ul Q rot g u e E u /kT ex e hν/kT ex − Z τ υ dυ (14)where A ul is the spontaneous emission coefficient fortransitions between levels u and l , g u is the degeneracyof level u , E u is the energy of level u , and T ex is theexcitation temperature. Q rot is the rotational partitionfunction, given by Q rot ≈ kT ex hB + 13 + 115 (cid:18) hB kT ex (cid:19) + 4315 (cid:18) hB kT ex (cid:19) + 1315 (cid:18) hB kT ex (cid:19) , (15)which is good to <
1% when T ex > B is the rota-tional constant of the molecule. τ ν is the optical depth of the line, and is related tobrightness temperature, T b , through the radiative trans-fer equation J ( T b ) = f [ J ( T ex ) − J ( T bg )][1 − e − τ v ] (16)where f is the beam filling factor and J ( T ) =( hν/k ) / ( e hν/kT − CO(J=1–0) is detected (case 1), we donot adopt the optically thick assumption but adopt theassumption that the excitation temperature of CO is 10K as suggested by Goldsmith et al. (2008) in the Tauruscloud. Optical depth τ v can be derived from equation 16.Then the total column density of CO can be derivedthrough combination of equations 14 and 15.For case 2, we adopted the assumption of optically thin CO(J=1–0) and CO(J=2–1) lines. This is reason-able since there is no CO(J=1–0) detection and thecorrected antenna temperature of CO are smaller than1 K (see e.g. source 3C105 in Fig. 13). The excitationtemperature of CO, T ex , is then obtained from the fol-lowing equation e . /T bg − . /T ex − e − . /T ex e . /T bg − . /T ex − ν A g ν A g × (cid:20) e . /T bg − e . /T bg − (cid:21) R T dυ R T dυ , (17)where T R = J ( T b ). Once T ex in equation 17 is deter-mined, N ( CO) is derived as that in case 1.When both CO(J=1–0) and CO(J=1–0) are de-tected as in case 3, we assume τ ( CO) ≫ T ex and τ ( CO) via: T ex = 5 . (cid:26) ln (cid:20) . T R ( CO) + 0 . (cid:21)(cid:27) − . (18) τ ( CO) = − ln { − T R ( CO)5 .
29 ( h e . / T ex − i − − . − } . (19) Li et al. N tot ( CO) is then obtained using equation 15,from which we compute N tot ( CO) by assuming[ C] / [ C]=68 (Milam et al. 2005).The central velocities of derived OH componentsare used as initial estimates for CO fitting with theCLASS/GILDAS software. The properties of the fittedCO components are shown in Table 3, and the abovecalculations of CO excitation temperature and total col-umn density are based on the derived Gaussian fittingresults. The coincidence between OH and CO compo-nents is judged by central velocity. The velocity differ-ence should be less than 0.5 km s − . We plot the relationbetween OH and CO column density in Figure 9. A pos-itive correlation between log(N(OH)) and log(N(CO)) isseen. Least-squares fitting yields the relation log(N(OH))= 8.16 . . + 0.316 . . log(N(CO)). The linear Pearsoncorrelation coefficient is 0.69, indicating a strong corre-lation. This result is consistent with Allen et al. (2015),where an apparent correlation was found between strong CO emission and OH emission.
Fig. 9.—
CO column densities, N (CO), versus OH column densi-ties, N (OH) for three categories of clouds. Molecular clouds (MG)with CO detections are represented with black filled circles in thelight yellow region. DMG-threshold clouds are represented by greyfilled circles in the dark yellow region. In these clouds, CO emis-sion is detected at the CO sensitivity of this work but would bedetected at less than 3 σ under a CO sensitivity of 0.25 K per 0.65km s − , typical of the CFA CO survey (Dame et al. 2001). Blacktriangles in the blue region represent DMG clouds in which COemission is not detected at the CO sensitivity of the present work.The red solid line represents the least-squares fit result to the MGand DMG-threshold clouds. CO-Dark Molecular Gas
All-sky CFA CO survey data (Dame et al. 2001) havebeen widely used to investigate the distribution of CO onlarge scales. Planck Collaboration et al. (2011), for ex-ample, analyzed Planck data along with the CFA CO sur-vey to probe the large-scale DMG distribution. In prin-ciple, the definition of a DMG cloud depends on the sen-sitivity of the CO data employed. For example, Donate& Magnani (2017) found that the fraction of CO-darkmolecular gas relative to total H decreased from 58% to30% in the Pegasus-Pisces region when higher sensitivityCO data were taken. In this study, the representative1- σ sensitivity of CFA CO data is about σ CFA =0.25 Kper 0.65 km s − . Using this as an illustrative detec- tion threshold, we here identify components as “DMG-threshold clouds” in cases where CO emission would beundetected at 3 σ CFA but is detected at the higher sensi-tivity of the present work.We compare the Gaussian components seen in H i ab-sorption, OH absorption, and CO emission. A total of219 H i and 49 OH absorption components were detected.Most OH components can be associated with an HI com-ponent within a velocity offset of 1.0 km s − , except forthree sources with offsets of about 1.5 km s − . Thereare four general categories of clouds, three of which areillustrated in Fig. 10. Toward 3C192, only H i is present,typical of CNM. Toward 3C133, H i , OH, and several COand CO isotopologues are all detected, which is repre-sentative of ‘normal’ molecular clouds. Toward 3C132,there exists a component with H i and OH absorption,but no CO emission, representing DMG. An example ofa DMG-threshold cloud can be found in the spectra of3C109 (Fig. 13), where only weak ( 0.1 K) CO is de-tected. In our sample, there are 77.6% CNM, 4.1% DMG,6.9% DMG-threshold and 11.4% molecular clouds. Interms of detectability in absorption, the apparent DMGclouds (DMG and DMG-threshold) are similar to molec-ular clouds with CO emission.The statistics of hydrogen column density are pre-sented in Figure 11. The hydrogen column density ofDMG clouds (DMG and DMG-threshold) falls betweenN(H) ∼ cm − and 4 × cm − corresponding toa extinction range A V ∼ iscomplete, while that of CO is not. All N (H) in this workrefer to hydrogen column density based on H i absorp-tion and CO emission measurements. Comparison with N (H ) based on Planck results will be published in a sep-arate paper (Hiep et al. in prep). The abundance of mea-sured CO is thus expected to be less than the canonicalvalue of 10 − and to vary with extinction. Our detectionstatistics of OH in DMGs are consistent with a picturein which gas in this intermediate extinction range is stillevolving chemically. OH is thus a potentially good tracerof diffuse gas with intermediate extinction, namely, be-tween the self-shielding thresholds of H and CO. Sucha suggestion has been borne out by detailed studies ofindividual regions with more complete information. Xu& Li (2016), for example, found much tighter X-factorsin both OH and CH than in CO, for gases in the inter-mediate extinction region of the Taurus molecular cloud. IMPLICATION OF ABSORPTION SURVEY
The expected locations and abundance of OH shouldmake it an excellent tracer of DMG. However, due bothto low optical depths and low contrast between the mainline excitation temperatures and the Galactic diffuse con-tinuum background, OH is generally difficult to detect inemission. Large-scale mapping of OH therefore requiresextremely high sensitivity observations and/or the exis-tence of bright continuum background against which thelines may be seen in absorption. A handful of studieshave accomplished the former over very limited regionsin the Northern Hemisphere local ISM (Barriault et al.2010; Allen et al. 2012, 2015; Cotten et al. 2012). TheSPLASH project (Dawson et al. 2014) has greatly im-proved on the sensitivities of previous large-scale surveysin the south, to map OH (primarily in absorption) overhere is OH and Does It Trace the Dark Molecular Gas (DMG)? 9 −10 0 10 20V lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−121.021.5 OH 1665 OH 1667
CNM −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.000.25 CO 1−0 CO 2−124.9025.05 OH 1665 OH 1667
020 HI
DMG −10 0 10 20V lsr (km s −1 )0.00.3 T A ( K ) CO 1−0 C O 1−0 03 CO 1−0 CO 2−12736 OH 1665 OH 1667
030 HI
Molecular Cloud
Fig. 10.—
Representative spectra as described in section 4.3. The 3C192 sightline has only H i seen in absorption. One component of the3C133 sightline has OH and H i in absorption and CO and its isotopologues in emission. The 3C132 sightline has one gas component withboth OH and H i , but no detectable CO transitions. Fig. 11.—
Histogram of hydrogen column density, N (H), forCNM (blue), DMG (filled gray), DMG-threshold (red), and molec-ular cloud (MG, green). N (H) contains contribution from H i forCNM, DMG, and DMG-threshold cloud. Both H i and H (trans-forming from CO measurements, N(H )=N(CO)/1 × − ) are in-cluded for MG cloud. The left and right dashed lines representN(H)= 9.4 × cm − (visual extinction A V = 0 .
05 mag) and3.8 × cm − (A V = 2 mag), respectively. >
150 square degrees of the bright inner Galaxy. How-ever, for outer-Galaxy and off-plane regions, the mostpractical approach will make use of upcoming radio tele-scopes to conduct comprehensive absorption surveys, ofthe kind piloted here.The Five-hundred-meter Aperture Spherical radioTelescope (FAST) commenced observing in September,2016. The unprecedented sensitivity of FAST and itsearly science instruments (Li et al. 2013) should makefeasible an H i +OH absorption survey, in the mode ofthe Millennium Survey, but with 10 times more sources.Figure 12 shows the distribution of potential continuumsources available to FAST. In the coming decades, theSKA1 will provide the survey speed and sensitivity tomeasure absorption with a source density of between a few to a few tens per square degree (McClure-Griffithset al. 2015). This makes feasible an all-sky “absorption-image”, mapping out a fine grid of ISM excitation tem-perature and column density over a very large fraction ofthe sky. Based on similar excitation and sensitivity con-siderations, ALMA is a powerful instrument to obtainsystematic and sensitive absorption measurements of mil-limeter lines in diffuse gas. CO and HCO + in diffuse gas,in particular, will be much better constrained in termsof excitation temperature and column densities throughALMA absorption observations than through emissionmeasurements. Combining both radio and millimeter ab-sorption surveys in the coming decade, we will quantifythe DMG and provide definitive answers to questions likethe global star formation efficiency. CONCLUSIONS
Utilizing unpublished OH absorption measurementsfrom the Millennium Survey and our own follow-up COsurveys, we carried out an analysis of the excitation con-ditions and quantity of OH along 44 sightlines throughthe Local ISM and Galactic Plane. CO was observed to-wards these positions. 49 OH components were detectedtowards 22 of these sightlines. The conclusions are asfollows:1. The excitation temperature of OH peaks around3.4 K and follows a modified normalized log-normaldistribution, f ( T ex ) ∝ √ πσ exp (cid:20) − [ln(T ex ) − ln(3 . σ (cid:21) . The majority of OH gas in our sample, presumablyrepresentative of the Milky Way, thus has an exci-tation temperature close to the background (CMBplus synchrotron), providing an explanation of whyOH has historically been so difficult to detect inemission.0 Li et al.
Fig. 12.—
Distribution of continuum point sources within the area of FAST sky coverage (limits shown with orange ), which covers adeclination range of -14.35 to 65.65 degrees. Red circles represent 372 sources with flux densities greater than 2.5 Jy in the NVSS survey.In initial observation periods, FAST will adopt a drift scan mode. The threshold of 2.5 Jy corresponds to a 3 σ detection in OH absorptionfor gas with an optical depth of 0.01, in a drift scan of 12 s, at a velocity resolution of 0.25 km s − , with a system temperature of 25 K.Blue filled circles represent 1071 sources with flux densities greater than 1.25 Jy in the NVSS survey. The threshold of 1.25 Jy allows fora 3 σ detection for OH having optical depth of 5.5 × − in a total observing time of 10 minutes (ON+OFF) in tracking mode. The greybackground is the integrated H i intensity map from the LAB H i survey (Hartmann & Burton 1997; Arnal et al. 2000; Bajaja et al. 2005).The limits of the coverage of Arecibo are shown with green solid lines. The positions of the 44 point sources used in this paper are plottedwith yellow filled squares.
2. The OH main lines are generally not in LTE, witha moderate excitation temperature difference of | T ex (1667) − T ex (1665) | < τ peaks at ∼ .
01, with the highest value inour sample equal to 0.22.4. A weak correlation between N (OH) and N (H i ) wasfound. The abundance ratio [OH]/[HI] has a me-dian of 10 − .5. N (OH) and N (CO) are linearly correlated whenboth are detected, which is consistent with previousobservations.6. Whether a cloud is designated as DMG depends onthe sensitivity of the CO data. By comparing withthe CfA CO survey data of Dame et al. (2001) wefind that the fraction of DMG components wouldincrease by a factor of ∼ V ∼ ACKNOWLEDGMENTS
This work is supported by National Key R&D Pro-gram of China 2017YFA0402600 and International Part-nership Program of Chinese Academy of Sciences No.here is OH and Does It Trace the Dark Molecular Gas (DMG)? 11114A11KYSB20160008. D.L. acknowledges supportfrom ”CAS Interdisciplinary Innovation Team” program.J.R.D. is the recipient of an Australian Research Coun-cil DECRA Fellowship (project number DE170101086).This work was carried out in part at the Jet PropulsionLaboratory, which is operated for NASA by the Califor-nia Institute of Technology. L.B. acknowledges supportfrom CONICYT Project PFB06 CO data were observed with the Delingha 13.7m telescope of the Qinghai Stationof Purple Mountain Observatory (PMODLH), the Cal-tech Submillimeter Observatory (CSO), and the IRAM30-meter telescope. The authors appreciate all the staffmembers of the PMODLH, CSO, and the IRAM 30-meter Observatory for their help during the observations.We thank Lei Qian and Lei Zhu for their help in CSOobservations.
REFERENCESAllen, R. J., Hogg, D. E., & Engelke, P. D. 2015, AJ, 149, 123Allen, R. J., Ivette Rodr´ıguez, M., Black, J. H., & Booth, R. S.2012, AJ, 143, 97Arnal, E. M., Bajaja, E., Larrarte, J. J., Morras, R., & P¨oppel,W. G. L. 2000, A&AS, 142, 35Bajaja, E., Arnal, E. M., Larrarte, J. J., et al. 2005, A&A, 440,767Barriault, L., Joncas, G., Lockman, F. J., & Martin, P. G. 2010,MNRAS, 407, 2645Bihr, S., Beuther, H., Ott, J., et al. 2015, A&A, 580, A112Boyce, P. J., & Cohen, R. J. 1994, A&AS, 107,Caswell, J. L., & Haynes, R. F. 1975, MNRAS, 173, 649Colgan, S. W. J., Salpeter, E. E., & Terzian, Y. 1989, ApJ, 336,231Cotten, D. L., Magnani, L., Wennerstrom, E. A., Douglas, K. A.,& Onello, J. S. 2012, AJ, 144, 163Crutcher, R. M. 1979, ApJ, 234, 881Crutcher, R. M. 1977, ApJ, 216, 308Dawson, J. R., Walsh, A. J., Jones, P. A., et al. 2014, MNRAS,439, 1596Dame, T. M., Hartmann, D., & Thaddeus, P. 2001, ApJ, 547, 792de Vries, H. W., Heithausen, A., & Thaddeus, P. 1987, ApJ, 319,723Destombes, J. L., Marliere, C., Baudry, A., & Brillet, J. 1977,A&A, 60, 55Dickey, J. M., Crovisier, J., & Kazes, I. 1981, A&A, 98, 271Donate , E., & Magnani, L. 2017, MNRAS, 472, 3169Field, G. B., Goldsmith, D. W., & Habing, H. J. 1969, ApJ, 155,L149Fukui, Y., Torii, K., Onishi, T., et al. 2015, ApJ, 798, 6Goldsmith, P. F., Heyer, M., Narayanan, G., et al. 2008, ApJ,680, 428-445Goss, W. M. 1968, ApJS, 15, 131Grenier, I. A., Casandjian, J.-M., & Terrier, R. 2005, Science,307, 1292Grossmann, V., Heithausen, A., Meyerdierks, H., & Mebold, U.1990, A&A, 240, 400Guibert, J., Rieu, N. Q., & Elitzur, M. 1978, A&A, 66, 395Harju, J., Winnberg, A., & Wouterloot, J. G. A. 2000, A&A, 353,1065Hartmann, D., & Burton, W. B. 1997, Atlas of Galactic NeutralHydrogen, by Dap Hartmann and W. Butler Burton,pp. 243. ISBN 0521471117. Cambridge, UK: CambridgeUniversity Press, February 1997., 243 Haslam, C. G. T., Salter, C. J., Stoffel, H., & Wilson, W. E. 1982,A&AS, 47, 1Heiles, C. 2001, ApJ, 551, L105Heiles, C., Perillat, P., Nolan, M., et al. 2001, PASP, 113, 1247Heiles, C., & Troland, T. H. 2003, ApJ, 586, 1067Heiles, C., & Troland, T. H. 2003, ApJS, 145, 329Langer, W. D., Velusamy, T., Pineda, J. L., et al. 2010, A&A,521, L17Langer, W. D., Velusamy, T., Pineda, J. L., Willacy, K., &Goldsmith, P. F. 2014, A&A, 561, A122Lee, M.-Y., Stanimirovi´c, S., Murray, C. E., Heiles, C., & Miller,J. 2015, ApJ, 809, 56Li, D., & Goldsmith, P. F. 2003, ApJ, 585, 823Li, D., Nan, R., & Pan, Z. 2013, Neutron Stars and Pulsars:Challenges and Opportunities after 80 years, 291, 325Liszt, H., & Lucas, R. 1996, A&A, 314, 917Lucas, R., & Liszt, H. 1996, A&A, 307, 237McClure-Griffiths, N. M., Stanimirovic, S., Murray, C., et al.2015, Advancing Astrophysics with the Square Kilometre Array(AASKA14), 130McKee, C. F., & Ostriker, J. P. 1977, ApJ, 218, 148Milam, S. N., Savage, C., Brewster, M. A., Ziurys, L. M., &Wyckoff, S. 2005, ApJ, 634, 1126Nguyen-Q-Rieu, Winnberg, A., Guibert, J., et al. 1976, A&A, 46,413Pineda, J. L., Langer, W. D., Velusamy, T., & Goldsmith, P. F.2013, A&A, 554, A103Pineda, J. L., Goldsmith, P. F., Chapman, N., et al. 2010, ApJ,721, 686Planck Collaboration, Ade, P. A. R., Aghanim, N., et al. 2011,A&A, 536, A19Sancisi, R., Goss, W. M., Anderson, C., Johansson, L. E. B., &Winnberg, A. 1974, A&A, 35, 445Stanimirovi´c, S., Murray, C. E., Lee, M.-Y., Heiles, C., & Miller,J. 2014, ApJ, 793, 132Turner, B. E. 1979, A&AS, 37, 1van Dishoeck, E. F., & Black, J. H. 1988, ApJ, 334, 771Wannier, P. G., Andersson, B.-G., Federman, S. R., et al. 1993,ApJ, 407, 163Weinreb, S., Barrett, A. H., Meeks, M. L., & Henry, J. C. 1963,Nature, 200, 829Wolfire, M. G., Hollenbach, D., & McKee, C. F. 2010, ApJ, 716,1191Wouterloot, J. G. A., & Habing, H. J. 1985, A&AS, 60, 43Xu, D., & Li, D. 2016, ApJ, 833, 90 −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.8 CO 1−0 CO 2−13031 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−129.429.7 OH 1665 OH 1667
030 HI −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.000.25 CO 1−0 CO 2−13334 OH 1665 OH 1667
030 HI −10 0 10 20V lsr (km s −1 )0.00.7 T A ( K ) CO 1−0 C O 1−006 CO 1−0 CO 2−1240260 OH 1665 OH 1667
Fig. 13.—
Spectra toward 3C105 (
T op left ), 3C109 ((
T op right )), 3C120 (
Bottom left ) and 3C123 (
Bottom right ).The Y axis, T A ,represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. here is OH and Does It Trace the Dark Molecular Gas (DMG)? 13 −10 0 10 20V lsr (km s −1 )0.00.2 T A ( K ) CO 1−0 C O 1−002 CO 1−0 CO 2−12224 OH 1665 OH 1667
030 HI −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−161.562.0 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−124.3024.45 OH 1665 OH 1667
030 HI −10 0 10 20V lsr (km s −1 )0.00.6 T A ( K ) CO 1−0 C O 1−0024 CO 1−0 CO 2−133.034.5 OH 1665 OH 1667
030 HI
Fig. 14.—
Spectra toward 3C131 (
T op left ), 3C138 ((
T op right )), 3C142.1 (
Bottom left ) and 3C154 (
Bottom right ). The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−0012 CO 1−0 CO 2−19.510.0 OH 1665 OH 1667
015 HI −20 −10 0 10V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.5 CO 1−0 CO 2−129.530.0 OH 1665 OH 1667 lsr (km s −1 )0 T A ( K ) CO 1−0 C O 1−00.00.6 CO 1−0 CO 2−120.020.8 OH 1665 OH 1667 lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−17.28.0 OH 1665 OH 1667
810 HI
Fig. 15.—
Spectra toward 3C167 (
T op left ), 3C18 ((
T op right )), 3C207 (
Bottom left ) and 3C225A (
Bottom right ). The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. here is OH and Does It Trace the Dark Molecular Gas (DMG)? 15 −10 0 10 20V lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−12022 OH 1665 OH 1667 lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−117.017.5 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−121.7522.00 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−0 CO 2−1160170 OH 1665 OH 1667
Fig. 16.—
Spectra toward 3C225B (
T op left ), 3C237 ((
T op right )), 3C245 (
Bottom left ) and 3C273 (
Bottom right ). The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. −20 −10 0 10V lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.000.45 CO 1−0 CO 2−112.7512.90 OH 1665 OH 1667 lsr (km s −1 )0.000.08 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−120.420.7 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−019.219.5 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−020.020.5 OH 1665 OH 1667
Fig. 17.—
Spectra toward 3C274.1 (
T op left ), 3C109 ((
T op right )), 3C315 (
Bottom left ) and 3C318 (
Bottom right ). The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. here is OH and Does It Trace the Dark Molecular Gas (DMG)? 17 −20 −10 0 10V lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−160.060.8 OH 1665 OH 1667 lsr (km s −1 )0.000.18 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−010.210.8 OH 1665 OH 1667
010 HI −10 0 10V lsr (km s −1 )0.000.18 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−0230240 OH 1665 OH 1667 lsr (km s −1 )0.00.2 T A ( K ) CO 1−0 C O 1−001 CO 1−0 CO 2−190.094.5 OH 1665 OH 1667
Fig. 18.—
Spectra toward 3C33 (
T op left ), 3C333 ((
T op right )), 3C348 (
Bottom left ) and 3C409 (
Bottom right ). The Y axis, T A ,represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. −10 0 10 20 30V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−001 CO 1−0 CO 2−17274 OH 1665 OH 1667 lsr (km s −1 )0.00.2 T A ( K ) CO 1−0 C O 1−00.00.8 CO 1−0 CO 2−1121122 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−114.114.4 OH 1665 OH 1667 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−001 CO 1−0 CO 2−118.819.2 OH 1665 OH 1667
Fig. 19.—
Spectra toward 3C410 (
T op left ), 3C454.3 ((
T op right )), 3C64 (
Bottom left ) and 3C75 (
Bottom right ). The Y axis, T A ,represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. here is OH and Does It Trace the Dark Molecular Gas (DMG)? 19 −20 −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−137.838.4 OH 1665 OH 1667
030 HI −10 0 10 20 30V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−137.538.0 OH 1665 OH 1667 lsr (km s −1 )0.00.6 T A ( K ) CO 1−0 C O 1−00.00.8 CO 1−017.518.0 OH 1665 OH 1667 lsr (km s −1 )0.01.2 T A ( K ) CO 1−0 C O 1−006 CO 1−0 CO 2−18.49.6 OH 1665 OH 1667
015 HI
Fig. 20.—
Spectra toward 3C78 (
T op left ), 3C98 ((
T op right )), 4C13.65 (
Bottom left ) and 4C13.67 (
Bottom right ). The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been corrected for themain-beam efficiency. −20 −10 0 10 20 30V lsr (km s −1 )0.00.3 T A ( K ) CO 1−0 C O 1−00 CO 1−0 CO 2−111.512.0 OH 1665 OH 1667
015 HI −10 0 10 20 30V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.4 CO 1−0 CO 2−115.615.9 OH 1665 OH 1667
G196.6+0.2 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−002 CO 1−0 CO 2−117.618.4 OH 1665 OH 1667
G197.0+1.1 lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.8 CO 1−0 CO 2−135.536.0 OH 1665 OH 1667
P0428+20
Fig. 21.—
Spectra toward 4C22.12 (
T op left ), G196.6+0.2 ((
T op right )), G197.0+1.1 (
Bottom left ) and P0428+20 (
Bottom right ). TheY axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been correctedfor the main-beam efficiency. here is OH and Does It Trace the Dark Molecular Gas (DMG)? 21 −20 −10 0 10 20V lsr (km s −1 )0.000.15 T A ( K ) CO 1−0 C O 1−00.00.3 CO 1−0 CO 2−15152 OH 1665 OH 1667
P0531+19 lsr (km s −1 )0.00.1 T A ( K ) CO 1−0 C O 1−00.00.2 CO 1−0 CO 2−111.412.0 OH 1665 OH 1667
P1055+20 lsr (km s −1 )0.00.5 T A ( K ) CO 1−0 C O 1−005 CO 1−0 CO 2−16.4 OH 1665 OH 1667
T0526+24
05 HI −20 −10 0 10 20 30 40 50V lsr (km s −1 )04 T A ( K ) CO 1−0 C O 1−005 CO 1−0 CO 2−11216 OH 1665 OH 1667
T0629+10
015 HI
Fig. 22.—
Spectra toward P0531+19 (
T op left ), P1055+20 ((
T op right )), T0526+24 (
Bottom left ) and T0629+10 (
Bottom right ).TheY axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightness temperature, which has been correctedfor the main-beam efficiency. −10 0 10 20V lsr (km s −1 )0.00.8 T b ( K ) CO 1−0 C O 1−00.01.2 CO 1−06000 OH 1665 OH 1667
TauA
Fig. 23.—
Spectra toward TauA. The Y axis, T A , represents antenna temperature, but for CO(1-0) spectra is the main-beam brightnesstemperature, which has been corrected for the main-beam efficiency. h e r e i s O H a nd D o e s I t T r a ce t h e D a r k M o l ec u l a r G a s ( D M G ) ? TABLE 2Gaussian Fit Parameters for OH main lines
Source l/b
OH(1665) OH(1667) τ V lsr ∆ V T ex N ( OH ) τ V lsr ∆ V T ex N ( OH )(Name) ( o ) ( kms − ) ( kms − ) ( K ) (10 cm − ) ( kms − ) ( kms − ) ( K ) (10 cm − )3C105 187.6/-33.6 0.0158 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L i e t a l. TABLE 3Gaussian Fit Parameters of CO Data
Source l/b O b N(12CO) T peaka Vlsr ∆ V T peaka V lsr ∆ V T peaka V lsr ∆ V (Name) ( ◦ ) K (km s −
1) (kms −
1) K (km s −
1) (km s − K (km s −
1) (km s −