Warm and cold molecular gas conditions modeled in 87 galaxies observed by the Herschel SPIRE FTS
Abstract
Molecular gas is the raw material for star formation, and like the interstellar medium (ISM) in general, it can exist in regions of higher and lower excitation. Rotational transitions of the CO molecule are bright and sensitive to cold molecular gas. While the majority of the molecular gas exists in the very cold component traced by CO J=1-0, the higher-J lines trace the highly excited gas that may be more indicative of star formation processes. The atmosphere is opaque to these lines, but the launch of the Herschel Space Observatory made them accessible for study of Galactic and extragalactic sources. We have conducted two-component, non-local thermodynamic equilibrium (non-LTE) modeling of the CO lines from J=1-0 through J=13-12 in 87 galaxies observed by the Herschel SPIRE Fourier Transform Spectrometer (FTS). We used the nested sampling algorithm Multinest to compare the measured CO spectral line energy distributions (SLEDs) to the ones produced by a custom version of the non-LTE code RADEX. This allowed us to fully examine the degeneracies in parameter space for kinetic temperature, molecular gas density, CO column density, and area filling factor.Here we discuss the major findings of our study, as well as the important implications of two-component molecular gas modeling. The average pressure of the warm gas is slightly correlated with galaxy LFIR, but that of the cold gas is not. A high-J (such as J=11-10) to J=1-0 line ratio is diagnostic of warm component pressure. We find a very large spread in our derived values of "alpha-CO," with no discernable trend with LFIR, and average molecular gas depletion times that decrease with LFIR. If only a few molecular lines are available in a galaxy's SLED, the limited ability to model only one component will change the results. A one-component fit often underestimates the flux of carbon monoxide (CO) J=1-0 and the mass. If low-J lines are not included, mass is underestimated by an order of magnitude. Even when modeling the low-J lines alone or using a CO-to-mass conversion factor, the mass should be considered to be uncertain to a factor of at least 0.4 dex, and the vast majority of the CO luminosity will be missed (median, 65 per cent).
- Publication:
-
American Astronomical Society Meeting Abstracts #231
- Pub Date:
- January 2018
- Bibcode:
- 2018AAS...23124709K