EPoS Contribution
EPoS Contribution
Pre-supernova feedback mechanisms drive the destruction of molecular clouds in nearby star-forming disc galaxies

Melanie Chevance
ITA/ZAH, Heidelberg, DE
Stellar feedback is a major driver of galaxy evolution, redistributing metals, energy and momentum from small scales to galactic scales. However, the exact physical processes responsible for stopping the accretion of gas onto giant molecular clouds (GMCs) and limiting the efficiency at which gas is converted into stars within these GMCs remain elusive. While feedback from supernova explosions has been the popular feedback mechanism included in simulations of galaxy formation and evolution, ‘early' feedback mechanisms such as stellar winds, photoionisation and radiation pressure are expected to play an important role in dispersing the gas after the onset of star formation. The unknowns about which feedback mechanism dominates at each step of the star formation process and what is the environment in which supernovae explode constitute major limitations of modern simulations of galaxy formation and evolution. I will present the first systematic characterisation of the duration over which GMCs are dispersed from homogeneous ALMA + optical observations at 50-100 pc resolution of a sample of star-forming disc galaxies. I will show that molecular clouds are rapidly destroyed by stellar feedback (within 1-5 Myr), which drastically limit their star formation efficiency to 2 to 10% depending on the galactic environment. Comparison of this rapid destruction timescale with analytical predictions demonstrates that, independently of the environment, pre-supernova feedback mechanisms (photoionisation and stellar winds) play a crucial role in dispersing GMCs. Finally, I will show that the efficiency at which these early feedback mechanisms couples with the parent GMC is relatively low (a few tens of per cent), such that the vast majority of momentum and energy emitted by the young stellar populations escapes the parent GMC. These measurements provide critical constrains to improve sub-grid recipes used in galaxy formation and evolution simulations.
Caption: Predicted feedback timescale versus observed feedback timescale for the nine galaxies in our sample. Within the black ellipse (bottom one), the predicted timescale is calculated as the minimum GMC dispersion time by stellar winds and photoionisation. All data points lie below the 1:1 relation (dashed line), implying that the coupling between stellar feedback and the surrounding molecular gas is not perfect. Within the grey ellipse (top one), the predicted timescale is calculated as the delay time of the first supernova explosion. In galaxies above the 1:1 dotted line, the feedback time-scale is smaller than the supernova delay time. In these galaxies, GMCs are destroyed predominantly by photoionisation and winds. In galaxies below the line, the feedback time-scale is longer than (but still consistent with) the supernova delay time. The comparison between the two set of predicted timescales suggests that GMCs are generally dispersed by early, pre-supernova feedback.
Collaborators:
D. Kruijssen, ARI/ZAH, DE
M. Krumholz, ANU, AU
B. Groves, UWA, AU
B. Keller, ARI/ZAH, DE
Key publication

Suggested Session: Molecular Clouds