We present a preliminary analysis of the data from the first two square degrees of the Milky Way Galactic Ring Survey (GRS), a ¹³CO and CS molecular line survey of the Inner Galaxy.

The ¹³CO velocity channel maps reveal an enormous amount of hitherto unresolved structure, partly due to the higher angular resolution and better sampling of the GRS compared to previous studies, but also due to the fact that the narrower line widths of the more optically thin ¹³CO lines allow a better separation of velocity components. The channel maps also demonstrate our ability to distinguish different overlapping clouds along the line of sight in velocity space. Our goal is to establish the kinematic distances to these sources.

Overlays of our new ¹³CO and CS data with an infrared image obtained with the MSX satellite at 8 micron wavelength reveal a striking correlation. Because we can associate infrared sources with molecular clumps, we can estimate their distances kinematically. We aim to find kinematic distances to a large number of faint young stars and protostars in order to establish their luminosity functions and spatial distributions throughout the 5 kpc Ring.

Using our new ¹³CO data together with the Boston University-Arecibo Galactic HI survey, we outline a method to solve the distance ambiguity. As an example, we present results for an extended ¹³CO cloud at 25 km s^-1 which coincides remarkably well with extinction on the Digital Sky Survey optical image. Its large extent and association with visual extinction as well as HI self-absorption places this filament at the near kinematic distance.

Kinematics

Fig.1: Maps of GRS 13CO 1-0 emission integrated  over 10 km s-1 wide velocity intervals. For each  subimage, the velocity interval is given in the upper  right corner. The plot range was chosen to start at  the average 3 sigma noise level of the observations.

We will use the high spectral resolution of our data in conjunction with standard clump finding algorithms to decompose the observed emission into clumps in configuration and velocity space. We will thus be able to derive fairly accurate positions of the clumps, their LSR velocities, and kinematic distances.

Here we focus on whether the infrared sources can be correlated with molecular clumps observed in ¹³CO and on how the kinematic distance ambiguity can be solved. To address these questions, we present an overlay of the GRS ¹³CO  first moment map with an infrared image obtained with the MSX satellite (Fig.2) and associate the extended molecular filament in the 25-30 km s^-1 velocity range with a feature seen in extinction in the optical and in self-absorption in HI (Fig.3 and 4).

Fig.2: Overlay of contours from the GRS ¹³CO 1-0 first moment map for emission between 0 and 80 km s^-1 on a false color infrared image obtained with the MSX satellite (Band A, 6.8-10.8 micron). The image has been background subtracted, which removes both Zodiacal and very large scale Galactic emission. The angular resolution of the infrared image is 36" compared with 46" of the GRS. Darker contours in ¹³CO correspond to intensity peaks. Positions from the IRAS point source catalog are indicated by blue stars.

The ¹³CO data show a remarkable correlation with the extended infrared emission. The correspondence of molecular clumps and compact infrared sources on small scales is even more striking. With the high resolution of both the GRS molecular line and modern infrared data (MSX, ISO, 2MASS), we will for the first time be able to associate a great number of embedded objects with molecular cloud fragments at distinct velocities. Our goal is ultimately to determine the distances to and luminosity function of the objects in this new sample.
 
 

Fig.3: Overlay of GRS ¹³CO 1-0 contours integrated over the velocity range 25-30 km s^-1 on a false color optical image of the appropriate region obtained from the Digital Sky Survey (DSS) provided by Skyview. The contours start at the average 3 sigma noise level of the map. Higher intensity contours are drawn in yellow to emphasize the locations of peak ¹³CO emission.

The morphology of the ¹³CO emission is a remarkable match to that of the dark regions of extinction seen in the optical image. Since peak ¹³CO emission, and hence high column density, is observed towards regions of higher optical extinction, we conclude that ¹³CO emission in this particular velocity range and the optical extinction originate in the same material.

The visual extinction on the DSS image is faint and might not be a useful indicator for nearby small clouds. A more promising approach to determine whether a cloud is near or far is to look for self-absorption (or the lack thereof) in a ubiquitous interstellar species: atomic hydrogen. Fig.4 shows a comparison of GRS ¹³CO and HI from the BU-Arecibo Galactic HI survey.
 
 
 

Fig.4: Overlay of GRS ¹³CO 1-0 contours on a false color atomic hydrogen image obtained for the BU-Arecibo Galactic HI survey (Bania 1999, private communication). The velocity range for integration in both cases is 25-30 km s^-1. The same filament that is seen in ¹³CO emission and in extinction at optical wavelengths is clearly seen in self-absorption in HI. Similar to the case of optical extinction, peak ¹³CO intensities correspond to maxima in HI self-absorption. Clearly cold HI is intimately associated with the molecular gas. These two independent results unambiguously place the ¹³CO emitting cloud at the near kinematic distance, i.e. 1.8 kpc from the Sun (using the Galactic rotation curve derived by Clemens 1985, ApJ, 295, 422). Note that this filament covers a large area on the sky, which is expected for a nearby cloud. Below, we outline our plans to solve the kinematic distance ambiguity. For this purpose, sample spectra of ¹³CO 1-0 and HI observed towards one selected position in the 25 km s^-1 cloud are shown in Fig.5.

The main objective of the Galactic Ring Survey is to study the distribution of molecular clouds and star forming regions in the Inner Milky Way at the highest angular resolution and finest sampling ever. With this unique database, we will be able to associate cloud cores with embedded infrared sources and clusters, and to derive their luminosity function.

This last step requires, however, that the distances to these objects be accurately known. The method to determine distances to embedded infrared sources is a two step procedure. First, the radial velocity of the Young Stellar Object's parental cloud core must be found. Second, a kinematic distance must be assigned.

The positions and radial velocities of clumps are provided by standard clump-finding algorithms applied to the ¹³CO data. We then determine whether an infrared source is present at the clump position and assign it the corresponding velocity. The observed radial velocities of molecular clumps and cloud fragments can be of course translated into kinematic distances if the Galactic rotation curve is known (e.g. Clemens 1985, ApJ, 295, 422).

While this gives a unique determination of the object's galactocentric radius, the line of sight distance to the Sun is intrinsically double-valued. This is a consequence of the geometry of the First Galactic Quadrant which gives the well known near-far kinematic distance ambiguity.

We can solve the distance ambiguity by using additional information supplied by modern 21cm HI observations. The BU-Arecibo Galactic HI survey (BUAO) provides the highest angular resolution (4') attainable by any single-dish radio telescopes at 21cm.

In the high resolution BUAO spectra, HI can be seen in self-absorption against background 21cm line emission if cold HI at the near kinematic distance absorbs 21cm radiation from a warm cloud at the far kinematic distance. As opposed to absorption observations toward a bright continuum source (e.g. HI and H₂CO), which require an HII region in the background, HI self-absorption is expected to be observable on a much larger scale since atomic hydrogen is a ubiquitous species in the interstellar medium and we expect to find a sufficient amount of cold residual HI within the molecular clouds of our study.

A striking example of an extended nearby cloud is shown in Fig.3 and 4. ¹³CO and HI spectra toward a selected position in this filament Fig.5) show a deep HI self-absorption dip at the radial velocity of ¹³CO emission (25 km s^-1). Besides this feature, many more velocity components along the same line of sight are visible. These correspond to smaller clouds and do not have equally obvious counterparts in the optical image. Only the presence (or absence) of HI self-absorption allows us to unambiguously assign the clouds to the near (or far) kinematic distance.
 
 

Fig.5: A comparison of GRS 13CO 1-0 and HI emission towards a  selected position in the 25 km s-1 filament. To enhance the signal  to noise and properly compare the data sets, the \drco\ data have  been resampled to the HI resolution and grid (4'). The presence  (or absence) of HI self-absorption at the 13CO velocity places the  cloud at the near (or far) kinematic distance.

Based on the initial survey results for the first two square degrees, we expect to identify several thousand clouds and cloud cores with the GRS. A great number of these objects will be associated with embedded infrared sources and clusters whose distances can now be estimated.

The GRS thus not only constitutes a unique database for determining the distribution of clouds, star forming regions, and the structure of the 5 kpc Ring, it also provides the key to establish for the first time the luminosity function of infrared sources and clusters throughout the Inner Milky Way.

A full size version of the poster can be obtained here (gzipped postscript file)

The GRS is supported by the NSF via grant AST-9800334 and AST-0098562