Measuring Star Formation Rates in the Milky Way from Hi-GAL 70 $\mu$m Observations

Abstract

Three methods for computing the total star formation rate of the Milky Way agree well with a reference value of $1.65\pm0.19$ M$_\odot$ yr$^{-1}$. They are then used to determine the radial dependence of the star formation rate and face-on map for the Milky Way. First, the method based on a model of star formation in Hi-GAL-defined dense clumps, adjusted for an increase in the gas-to-dust ratio with Galactocentric radius, predicts $1.65\pm0.61$ M$_\odot$ yr$^{-1}$. Second, the method using the 70 $\mu$m emission, commonly used in other galaxies, with a technique to assign distances to the extended emission, predicts $1.42^{+0.63}_{-0.44}$ M$_\odot$ yr$^{-1}$. Finally, a method based on theoretical predictions of star formation efficiency as a function of virial parameter, with masses corrected for metallicity dependence, applied to a catalog of molecular clouds also predicts a value in agreement at $1.47$ M$_\odot$ yr$^{-1}$. The three methods predict the radial variation of the star formation rate, with remarkably good agreement from the CMZ out to about 20 kpc. More differences were seen in face-on maps with a resolution of 0.5 kpc made with the three approaches and in comparisons to the local (within 3 kpc) star formation rate, indicating limitations of the methods when applied to smaller scales. The 70 $\mu$m star formation rate follows very closely the surface density of molecular gas, corrected for a metallicity-dependent CO conversion factor. A molecular gas depletion time of 1 Gyr is consistent with the data, as is a molecular Kennicutt-Schmidt relation with a power-law slope of $1.10 \pm 0.06$.