Based on sensitive CO measurements from HERACLES and HI data from THINGS, we show that the azimuthally averaged radial distribution of the neutral gas surface density (Sigma_HI + Sigma_H2) in 33 nearby spiral galaxies exhibits a well-constrained univ
ersal exponential distribution beyond 0.2*r25 (inside of which the scatter is large) with less than a factor of two scatter out to two optical radii r25. Scaling the radius to r25 and the total gas surface density to the surface density at the transition radius, i.e., where Sigma_HI and Sigma_H2 are equal, as well as removing galaxies that are interacting with their environment, yields a tightly constrained exponential fit with average scale length 0.61+-0.06 r25. In this case, the scatter reduces to less than 40% across the optical disks (and remains below a factor of two at larger radii). We show that the tight exponential distribution of neutral gas implies that the total neutral gas mass of nearby disk galaxies depends primarily on the size of the stellar disk (influenced to some degree by the great variability of Sigma_H2 inside 0.2*r25). The derived prescription predicts the total gas mass in our sub-sample of 17 non-interacting disk galaxies to within a factor of two. Given the short timescale over which star formation depletes the H2 content of these galaxies and the large range of r25 in our sample, there appears to be some mechanism leading to these largely self-similar radial gas distributions in nearby disk galaxies.
We present a detailed analysis comparing the velocity fields in molecular clouds and the atomic gas that surrounds them in order to address the origin of the gradients. To that end, we present first-moment intensity-weighted velocity maps of the mole
cular clouds and surrounding atomic gas. The maps are made from high-resolution 13CO observations and 21-cm observations from the Leiden/Argentine/Bonn Galactic HI Survey. We find that (i) the atomic gas associated with each molecular cloud has a substantial velocity gradient---ranging within 0.02 to 0.07 km/s/pc---whether or not the molecular cloud itself has a substantial linear gradient (ii) If the gradients in the molecular and atomic gas were due to rotation, this would imply that the molecular clouds have less specific angular momentum than the surrounding HI by a factor of 1-6. (iii) Most importantly, the velocity gradient position angles in the molecular and atomic gas are generally widely separated---by as much as 130 degrees in the case of the Rosette Molecular Cloud. This result argues against the hypothesis that molecular clouds formed by simple top-down collapse from atomic gas.
We present an analysis comparing the properties of 45 giant molecular clouds (GMCs) in M33 and the atomic hydrogen (HI) with which they are associated. High-resolution VLA observations are used to measure the properties of HI in the vicinity of GMCs
and in regions where GMCs have not been detected. The majority of molecular clouds coincide with a local peak in the surface density of atomic gas, though 7% of GMCs in the sample are not associated with high-surface density atomic gas. The mean HI surface density in the vicinity of GMCs is 10 M_sol/pc^2 and tends to increase with GMC mass as Sigma_HI ~ M_GMC^0.27. 39 of the 45 HI regions surrounding GMCs have linear velocity gradients of ~0.05 km/s/pc. If the linear gradients previously observed in the GMCs result from rotation, then 53% are counterrotating with respect to the local HI. If the linear gradients in these local HI regions are also from rotation, 62% are counterrotating with respect to the galaxy. If magnetic braking reduced the angular momentum of GMCs early in their evolution, the angular velocity of GMCs would be roughly one order of magnitude lower than what is observed. Based on our observations, we consider the possibility that GMCs may not be rotating. Atomic gas not associated with GMCs has gradients closer to 0.03 km/s/pc, suggesting that events occur during the course of GMC evolution that may increase the shear in the atomic gas.
