Online material

Appendix A: Errors and uncertainties

As conventional error propagation breaks down when the uncertainties become larger than a few percent, one can only point out the individual sources of errors and estimate the final uncertainties.

The distance error mainly stems from uncertainties in the Galactic rotation curve and errors in the gas velocities can be neglected. Error propagation including the velocity uncertainties following Reid et al. (2009) suggests uncertainties of smaller than 0.1 kpc. However, owing to intrinsic errors and deviations from the global Galactic rotation we estimate the distance to be uncertain to within 0.5 kpc. This leads to a contribution to the final mass uncertainties ranging from 10% to 50%, that depends on the absolute distance. In addition, individual objects close to the Galactic center that have non-circular orbits may be placed at random distances and contaminate the sample.

In the literature, temperatures of starless cores range from 10 K to 20 K with the bulk at 15 K (Sridharan et al. 2005; Pillai et al. 2006; Vasyunina et al. 2011; Peretto et al. 2010). The temperature estimate here is based on direct observations of 15 out of 210 starless clumps and is in good agreement with earlier studies. A temperature uncertainty of  ±5 K at 15 K may introduce mass uncertainties of about a factor of two.

The dust properties and the gas-to-dust ratio are very uncertain as well and might contribute another factor of two to the errors. The flux uncertainties are dominated by the calibration uncertainties, which are  ~15% (Schuller et al. 2009).

When calculating the column density as well as the masses, the predominant uncertainties are those of the dust properties and temperatures. For the mass, the uncertainty in the distance is equally important. Altogether, the total uncertainties in the mass may be as large as a factor of five.

Appendix B: Omitted regions.

Appendix C: Stamps of starless regions

    

© ESO, 2012