A&A 518, L62 (2010)

Herschel: the first science highlights

LETTER TO THE EDITOR

Mapping the interstellar medium in galaxies with Herschel/SPIRE

S. A. Eales¹ - M. W. L. Smith¹ - C. D. Wilson¹⁶ - G. J. Bendo³ - L. Cortese¹ - M. Pohlen¹ - A. Boselli⁴ - H. L. Gomez¹ - R. Auld¹ - M. Baes² - M. J. Barlow⁵ - J. J. Bock⁶ - M. Bradford⁶ - V. Buat⁴ - N. Castro-Rodríguez⁷ - P. Chanial⁸ - S. Charlot⁹ - L. Ciesla⁴ - D. L. Clements³ - A. Cooray¹⁰ - D. Cormier⁸ - J. I. Davies¹ - E. Dwek¹¹ - D. Elbaz⁸ - M. Galametz⁸ - F. Galliano⁸ - W. K. Gear¹ - J. Glenn¹² - M. Griffin¹ - S. Hony⁸ - K. G. Isaak¹³ - L. R. Levenson⁶ - N. Lu⁷ - S. Madden⁸ - B. O'Halloran³ - K. Okumura⁸ - S. Oliver¹⁴ - M. J. Page¹⁵ - P. Panuzzo⁸ - A. Papageorgiou¹ - T. J. Parkin¹⁶ - I. Pérez-Fournon⁷ - N. Rangwala¹² - E. E. Rigby¹⁷ - H. Roussel⁹ - A. Rykala¹ - N. Sacchi¹⁸ - M. Sauvage⁸ - B. Schulz¹⁹ - M.R. P. Schirm¹⁶ - L. Spinoglio¹⁸ - S. Srinivasan⁹ - J. A. Stevens²⁰ - M. Symeonidis¹⁵ - M. Trichas³ - M. Vaccari²¹ - L. Vigroux⁹ - H. Wozniak²² - G. S. Wright²³ - W. W. Zeilinger²⁴

1 - School of Physics and Astronomy, Cardiff University, The Parade, Cardiff CF24 3AA, UK
2 - Sterrenkundig Observatorium, Universiteit Gent, Krijgslaan 281 S9, 9000 Gent, Belgium
3 - Astrophysics Group, Imperial College, Blackett Laboratory, Prince Consort Road, London SW7 2AZ, UK
4 - Laboratoire d'Astrophysique de Marseille, UMR6110 CNRS, 38 rue F. Joliot-Curie, 13388 Marseille France
5 - Dept. of Physics and Astronomy, University College London, Gower Street, London WC1E 6BT, UK
6 - Jet Propulsion Laboratory, Pasadena, CA 91109, California Institute of Technology, Pasadena, CA 91125, USA
7 - Instituto de Astrofísica de Canarias (IAC) and Departmento de Astrofísica, Universidad de la Laguna (ULL), La Laguna, Tenerife, Spain
8 - CEA, Laboratoire AIM, Irfu/SAp, Orme des Merisiers, 91191 Gif-sur-Yvette, France
9 - Institut d'Astrophysique de Paris, UMR7095 CNRS, 98bis boulevard Arago, 75014 Paris, France
10 - Dept. of Physics & Astronomy, University of California, Irvine, CA 92697, USA
11 - Observational Cosmology Lab, Code 665, NASA Goddard Space Flight Center Greenbelt, MD 20771, USA
12 - Department of Astrophysical and Planetary Sciences, CASA CB-389, University of Colorado, Boulder, CO 80309, US
13 - ESA Astrophysics Missions Division, ESTEC, PO Box 299, 2200 AG Noordwijk, The Netherlands
14 - Astronomy Centre, Department of Physics and Astronomy, University of Sussex, UK
15 - Mullard Space Science Laboratory, University College London, Holmbury St Mary, Dorking, Surrey RH5 6NT, UK
16 - Dept. of Physics & Astronomy, McMaster University, Hamilton, Ontario, L8S 4M1, Canada
17 - School of Physics & Astronomy, University of Nottingham, University Park, Nottingham NG7 2RD, UK
18 - Istituto di Fisica dello Spazio Interplanetario, INAF, Via del Fosso del Cavaliere 100, 00133 Roma, Italy
19 - Infrared Processing and Analysis Center, California Institute of Technology, 770 South Wilson Av, Pasadena, CA 91125, USA
20 - Centre for Astrophysics Research, Science and Technology Research Centre, University of Hertfordshire, Herts AL10 9AB, UK
21 - University of Padova, Department of Astronomy, Vicolo Osservatorio 3, 35122 Padova, Italy
22 - Observatoire Astronomique de Strasbourg, UMR 7550 Université de Strasbourg - CNRS, 11 rue de l'Université, 67000 Strasbourg, France
23 - UK Astronomy Technology Center, Royal Observatory Edinburgh, Edinburgh EH9 3HJ, UK
24 - Institut für Astronomie, Universitt Wien, Trkenschanzstr. 17, 1180 Wien, Austria

Received 29 March 2010 / Accepted 19 April 2010

Abstract
The standard method of mapping the interstellar medium in a galaxy, by observing the molecular gas in the CO 1-0 line and the atomic gas in the 21-cm line, is largely limited with current telescopes to galaxies in the nearby universe. In this letter, we use SPIRE observations of the galaxies M 99 and M 100 to explore the alternative approach of mapping the interstellar medium using the continuum emission from the dust. We have compared the methods by measuring the relationship between the star-formation rate and the surface density of gas in the galaxies using both methods. We find the two methods give relationships with a similar dispersion, confirming that observing the continuum emission from the dust is a promising method of mapping the interstellar medium in galaxies.

Key words: galaxies: ISM - galaxies: spiral - dust, extinction - ISM: abundances

1 Introduction

Two of the key measurements of a galaxy that are needed for any investigation of galactic evolution are the star formation rate and, because stars form out of gas, the mass of interstellar material. The standard method of measuring the gas mass is to use the 21-cm line to measure the mass of the atomic gas and the CO 1-0 line to measure the mass of the molecular material. The conversion between 21-cm flux and the mass of atomic gas is unambiguous, but the conversion factor between the CO luminosity and the mass of molecular gas (the notorious ``X factor'') is uncertain both because the CO 1-0 line is optically thick and because the CO molecule is only tracing the much larger number of unseen hydrogen molecules (e.g. Bell et al. 2007). Both techniques suffer from the problem that with current telescopes it is very difficult to measure gas masses for galaxies beyond a redshift of 0.2, a major limitation now that the Herschel Space Observatory (Pilbratt et al. 2010) is detecting thousands of galaxies in the high-redshift universe (e.g. Eales et al. 2010).

The idea of using the continuum emission from the dust as an alternative way of mapping the interstellar medium (ISM) has a long history, dating back at least to Hildebrand (1983). More recently, Guelin et al. (1993, 1995) and Boselli et al. (2002) have attempted to determine the X-factor using the continuum emission from the dust and the assumption that the gas-to-dust ratio is the same in both the atomic and molecular phase of the ISM. The problem in attempting to measure the mass of the ISM from the continuum emission of the dust is that the conversion factors between the submillimetre flux and the mass of dust and between the dust mass and the gas mass are both poorly known. These constants have traditionally been estimated from galactic objects, such as reflection nebulae (e.g. Hildebrand 1983), but James et al. (2002) have argued that a better approach is to estimate them from the galaxies themselves, which is the approach we follow here.

In this letter, we have used both methods to map the ISM in two nearby galaxies, M 99 and M 100, which were two of the first targets to be observed in the Herschel Reference Survey (Boselli et al. 2010), a survey with SPIRE (Griffin et al. 2010) of 323 galaxies in a volume-limited sample of the local universe with distances between 15 and 25 Mpc. To assess the accuracy of the methods, we have used these maps to investigate the relationship between the star-formation rate in these galaxies and the density of the gas out of which the stars form. This is usually paramaterised by a power-law relationship between the star-formation rate per unit volume and the gas density, (Schmidt 1959), which for a constant scale height will result in a relationship between the star-formation rate per unit area of the disk and the surface density of the ISM which is also a power law with the same index: . This relationship is of great interest since different theoretical models predict values of N between 0.75 and 2 (Bigiel et al. 2008 and references therein), but in this paper we are less interested in the value of N than in using the fact that there is undoubtedly some intrinsic relationship between and to assess the errors, both statistical and systematic, in both methods for estimating the surface-density of the ISM.

2 Mapping the ISM using emission from the dust

The method of James et al. (2002) is based on the assumption that a constant fraction of metals in the ISM is incorporated in dust grains. The mass of the ISM in the galaxy is then given by:

in which f is the ratio of the mass of metals to the mass of hydrogen in a gas with solar abundance, is the ratio of the mass of metals in the dust to the total mass of metals, Z is the metallicity of the galaxy in units of solar metallicity, and is the mass-opacity coefficient of the dust at the frequency at which the flux has been measured. If the frequency is low enough, the Planck function reduces to the Rayleigh-Jeans approximation and the gas mass depends only linearly on the dust temperature. The importance of SPIRE is that its long-wavelength channels (350 and 500 m) mean that it is much more likely this approximation is true than for previous space observatories such as IRAS, ISO and Spitzer.

James et al. (2002) used abundance measurements of the local ISM to estimate a value for of 0.46. We have used a value for f of 0.019 and a solar metal abundance of (Asplund et al. 2009). The dust temperature can be estimated from the spectral energy distribution of the galaxy and Z can be estimated from optical spectroscopy. The only unknowns are therefore and , and so in principle if one has a single galaxy for which one knows the gas mass one can estimate  . On the assumption that the properties of the dust and the fraction of metals in the dust do not vary from galaxy to galaxy, one can then use this method to estimate the mass of the interstellar medium in a galaxy for which there is not a direct measurement of the gas mass.

3 The data

M 99 (NGC 4254) and M 100 (NGC 4321) are two spiral galaxies in the Virgo Cluster, which we take to have a distance of 16.8 Mpc. We have used the maps of the star-formation rate in these galaxies produced by Wilson et al. (2009), who combined 24-m Spitzer images and H images to correct for the effect of dust obscuration, following the prescription of Calzetti et al. (2007). We have produced maps of the gas in each galaxy by combining the CO 1-0 maps of Kuno et al. (2007) and the HI maps from the VIVA survey (Chung et al. 2009). Before combining the maps, we convolved the CO map to the same resolution as the 350-m map (FWHM of 25 arcsec). The HI map has slightly worse resolution than the 350-m data and so we did not smooth it. In estimating the mass of molecular material from the CO 1-0 line, we have used an X factor of (Bolatto et al. 2008). In M 99, the molecular material dominates the atomic gas within 80 arcsec of the galaxy centre and in M 100 within 100 arcsec.

We observed M 99 and M 100 during the Herschel Science Demonstration Phase with SPIRE. The calibration methods and accuracy of SPIRE are described by Swinyward et al. (2010). Images of the galaxies at the three SPIRE wavelengths (250, 350 and 500 m) are shown in Pohlen et al. (2010). To estimate the distribution of dust temperature in each galaxy, which is necessary to estimate the surface-density of the ISM (Eq. (1)), we used the SPIRE images at 250 and 350 m and the Spitzer images at 70 m (the SPIRE image at 500 m does not have sufficient angular resolution). We convolved the two short-wavelength maps to the resolution of the 350-m image and rebinned the images with a pixel size of 10 arcsec. At each pixel, we fitted a single-temperature grey-body ( ) with an emissivity index, , of 2 to the three fluxes to estimate the dust temperature. The temperatures we derive are in the range 17 < T < 25 K. To determine the metallicity, we used the radial profiles in Skillman et al. (1996). The remaining unknown on the right of Eq. (1) is . We estimated its value at 350 m by scaling, using , from the value of 0.07 measured at 850 m by James et al., giving a value of 0.41 . This is highly uncertain but, as we will show below, it is possible in principle to recalibrate using the observations themselves. The flux density we used in Eq. (1) was the measured 350-m flux density. All the maps were rebinned into images with the same pixel size (10 arcsec) and astrometrically registered.

Table 1:   Fits.

4 The star-formation rates in M 99 and M 100

Figures 1 and 2 show the star-formation rate per unit area plotted against the surface density of the ISM for M 99 (Fig. 1) and for M 100 (Fig. 2). We have binned the images into 30-arcsec pixels before plotting these diagrams, so that the pixels are bigger than the beam size and the data for each point in the figures are clearly independent. At the distance of Virgo, a 30-arcsec pixel corresponds to 2.4 kpc. The left-hand panel in each figure shows the result from the ISM map produced from the CO and HI data, and the right-hand panel shows the result from using the ISM map produced from our alternative method.

The two panels look qualitatively very similar for both galaxies, with a tight relationship in the region of each galaxy where the ISM is dominated by molecular material (the open circles) and a much weaker relationship outside this region. There is a significant offset between the two relationships along the abscissa, which must be due to systematic errors in one or both of the methods (Sect. 5). For both methods and galaxies, we used the bisector method (Isobe et al. 1990) to fit a straight line to the points within the region dominated by molecular gas. On the assumption that there is an intrinsic relationship between the star-formation rate and the surface-density of the ISM, one way to compare the accuracy of the two methods of mapping the ISM is to compare the residuals around the best-fit lines. Table 1 shows the root-mean-square residuals of the points around each line along both the ordinate and the abcissa for both methods. Based on this criterion, mapping the ISM from the dust emission works best for M 100 but mapping the ISM using the gas observations works best for M 99.

There are, however, clearly systematic errors in one or both methods because the values of N determined from the two methods differ by more than expected from the errors on each value. A simple test shows that this is not necessarily the result of a problem with the dust method. We have repeated the analysis using a value for the X-factor of , the value suggested by Nakai and Kuno (1995). Restricting the analysis to the region in each galaxy in which the molecular gas dominates, we find for M 99 and for M 100. The large changes show the different values of N may be as easily caused by systematic errors in the X factor as in our alternative method.

Although the focus of this paper is on testing methods for mapping the ISM in galaxies, we note that our results on the relationship between star-formation rate and the surface density agree well with recent results. Kennicutt et al. (2007) investigated this relationship within M 51 and Bigiel et al. (2008) investigated it in 18 nearby galaxies. Both studies concluded that there was little relationship between star-formation rate and the surface-density of atomic gas but a strong one between the star-formation rate and the surface-density of molecular gas. This is in qualitative agreement with our finding of a strong correlation in the inner part of each galaxy, where the molecular gas is dominant. Bigiel et al. (2008) found that if one only considers the surface-density of molecular gas the value of N is 1. Wilson et al. (2009) used CO 3-2 observations to map the ISM in the same galaxies we have considered in this paper and found little variation in the gas depletion timescale across the disks, which implies . Our values of N are higher than this, whichever method we use for mapping the ISM. However, if we only use the CO 1-0 observations to map the gas, we obtain for M 99 and for M 100, suggesting that it is the atomic phase of the ISM in the central parts of these galaxies which is responsible for the higher values of N.

5 Calibrating the mass-opacity coefficient

We have attempted to calibrate by requiring that the best-fit lines in the left-hand and right-hand panels of Figs. 1 and 2 agree at , a value chosen because it corresponds to the region in which the relationship between the gas surface-density and the star-formation density is tightest. We obtain a value for at 350 m of 0.056  for M 99 and 0.063 for M 100.

These values are three times lower than the estimate from the theoretical models of Li and Draine (2001) of . The only previous attempt to determine the submillimetre opacity at this wavelength directly from observations was that of Boulanger et al. (1996), who used COBE observations of dust at high galactic latitude and HI observations to obtain a relationship between submillimetre optical depth and hydrogen column density. Assuming solar metallicity and the same values of f and used in this paper (assumptions which are equivalent to assuming a gas-to-dust ratio of 153 for the high-latitude gas), we derive a value for of 0.35 , again higher than our value. Our value is also, of course, significantly lower than the value (0.41  ) obtained by extrapolating the measurement at 850 m of James et al. (2002) using a dust-emissivity index of 2.

The discrepancy between our result and the James et al. result is surprising, since the method is the same in both cases, although James et al. estimated from integrated gas and dust measurements for a large sample of galaxies whereas we have estimated it from maps of the gas and dust for indvidual galaxies. Since our result depends critically on the assumed value of the X factor, one way to bring the results into line would be if the values of the X factor for M 99 and M 100 are 6 times lower than the value we have assumed, which seems unlikely, however, given the recent evidence that the X factor is fairly constant between galaxies (Bolatto et al. 2008). It seems more plausible that at least part of the discrepancy is caused by our over-estimating the temperature of the dust. When fitting single-temperature models, both non-equilibrium heating of grains and the simple fact that warm dust produces more radiation than cold dust (Eales et al. 1989) can bias the fits towards higher temperatures. The temperatures produced by our fits lie in the range 17 < T < 25 K and the lack of any strong dependence of dust temperature on star-formation rate agrees with the recent Herschel study of the much closer galaxy M81 (Bendo et al. 2010). The temperatures are not much higher than the temperature of high-latitude galactic dust measured by Boulanger et al. (1996): 17.5 K. Nevertheless, if the temperature of the dust in both M 99 and M 100 were actually 10 K rather than 20 K,this would be enough to increase the value of to the value expected from the James et al. study. There are other possible problems with this method, such as disagreements between different methods for measuring metallicity (Kewley & Ellison 2008) and variations in and both within and between galaxies, but the biggest practical problem at present is the difficulty of accurately measuring the temperature of the dust. Fortunately, this should be simple to overcome by combining the Herschel observations with observations at longer wavelengths, for example at 850 m with LABOCA and SCUBA-2.

In summary, the remarkable similarity between the right-hand and left-hand panels of Figs. 1 and 2 shows that observing the continuum emission from dust is a promising method of mapping the interstellar medium in galaxies. Nevertheless, it is clear that considerable work needs to be done to calibrate this method, both in determining more accurately the temperature of the dust in individual galaxies and in determing the variation in and the other properties of the dust in the galaxy population. The Herschel Reference Survey and the other Herschel surveys of nearby galaxies are ideal for this kind of study.

Acknowledgements

SPIRE has been developed by a consortium of institutes led by Cardiff Univ. (UK) and including Univ. Lethbridge (Canada); NAOC (China); CEA, LAM (France); IFSI, Univ. Padua (Italy); IAC (Spain); Stockholm Observatory (Sweden); Imperial College London, RAL, UCL-MSSL, UKATC, Univ. Sussex (UK); Caltech, JPL, NHSC, Univ. Colorado (USA). This development has been supported by national funding agencies: CSA (Canada); NAOC (China); CEA, CNES, CNRS (France); ASI (Italy); MCINN (Spain); Stockholm Observatory (Sweden); STFC (UK); and NASA (USA).

References

Footnotes

...Herschel

Herschel is an ESA space observatory with science instruments provided by European-led Principal Investigator consortia and with important participation from NASA.

All Tables

Table 1:   Fits.

All Figures

---

Copyright ESO 2010