PowerPoint Presentation

The Interstellar Mediumand Star Formation Processin the Andromeda GalaxyAndreas SchrubaMax-Planck-Institut fr extraterrestrische PhysikESO 3D2014: Gas and Stars in GalaxiesMarch 10, 20141

ESO 3D2014: Gas and Stars in GalaxiesMarch 10, 2014

The CARMA Survey of Andromeda

Todays largest (0.1 deg2) Interferometric Survey ofCO(1-0) on Cloud Scale (20 pc x 1 km/s)Good Morning and Hello Everyone!

I would like to tell you a little more about the star formation process in nearby galaxies, but in contrary to Adams fabulous talk on starbursts, I want to speak about normal SF galaxies namely the Andromeda

In particular I want to show you first results from the CARMA Survey of Andromeda which is todays largest interferometric survey of CO emission in an external galaxy... and provides a high resolution view on individual molecular clouds2

Is there Concordance of Cloud Properties?3What are the Properties of Molecular Clouds?2Where is the Molecular Gas in a normal Spiral?1

ESO 3D2014: Gas and Stars in GalaxiesMarch 10, 2014So I want to use these data to ask the following three questions:

Where is the molecular gas in a normal spiral in GMCs or not?What are the properties of molecular clouds are they universal?Is there concordance of cloud properties for neighboring cloud?3Where is the Molecular Gas in a Spiral?1

Schruba+ (in prep)

a 3D cloud-scale view of CO(1-0) in AndromedaBut before I begin, let me offer you this truly 3D view on the H2 gas... this video runs through frequency space and each little structure here is a individual molecular cloud, the smallest ones not much bigger than Taurus (our point source sensitivity limit).4 a 2D cloud-scale view of CO(1-0) in AndromedaWhere is the Molecular Gas in a Spiral?1

Schruba+ (in prep)

Peak BrightnessLet me end this video by showing you also the peak brightness distribution which really nicely brings out all the detailed structures thought unfortunately, it seems that this screen is a little to small to really show all the information there is not sure what the architect of beautiful building was thinking here 5Where is the Molecular Gas in a Spiral?1 PDFs of CO Pixel BrightnessHughes+13; Gratier+12Data at Common Resolution of 50 pc x 5 km/s

Log width ~ 0.26 dexM31

~ 0.23 dexLMC

~ 0.33 dexM51Most emission at low intensity (signal-to-noise) and thus often missed.To start characterizing where the CO signal comes from, I show you the PDF of pixel brightness (a really poor tracer of volume density)Since the data cube contains a lot of signal-free regions, I tried to remove the noise contribution hereTo first order the distribution follows a log-normal w/ width of 0.26

For comparison I also want to show you the PDFs for LMC & M51 while the one for the LMC is quite comparable, the one for M51 is different, it is wider and reached much higher brightness indicating that strong spiral potential or active star formation may shift the distributions

However, the peak of the distribution is poorly defined, and thus the bulk of the flux is at low intensities within the noise and easily missed

6A Thick Disk of (diffuse) Molecular Gas1Caldu-Primo+13; Pety+13

Velocity Dispersion of HI and CO in 12 spirals at low inclinationLine Width of HI and CO comparable=> Evidence for thick disk of diffuse H2Caldu-Primo+ 13M31: cumulative emissionaround local HI velocity

M3176% of flux within ~40 km/sLOS dispersion ~18 km/s but M31 highly inclinedTo test where the low intensity CO signal is coming from, I integrate the CO flux within a velocity window around the local mean gas velocity and then vary the width of the window

First of all the curves for CO and HI are not vastly different.... for the CO, ~76% of all flux is contained within 40 km/s corresponding to a line-of-sight dispersion of 18 km/s.

But the velocity dispersions are clearly easier measured in more face-on galaxies, as has been done recently for 12 nearby spirals by Anahi Caldu-Primo. She found basically identical velocity dispersions for HI and CO around 12 km/s. This provides strong evidence there atomic and molecular gas are well mixed and that there must be a diffuse, thick molecular gas disk.7Mass Spectrum of Molecular Clouds2Colombo+13; Gratier+12; Wong+11; Rosolowsky+05; Heyer+01

10-kpc ring: = -2.05-kpc ring: = -1.4M31 as described by truncated power-law

M51All data: = -2.3range: -1.3 -2.6

M33All data: = -2.0range: -1.5 -2.3Cloud mass spectrum steeper in low-density then high-density environments, but cloud samples contain only ~50% of total mass, rest in diffuse phase10-kpc ring5-kpc ringBut lets look at the properties of mol clouds, eg the mass spectrum which is typically described by (truncated) power-lawsMost of our survey targets the 10-kpc gas ring for which we find a power-law index of -2, eg equal mass in low- and high-mass clouds but in the inner 5-kpc ring, the spectrum is significantly shallower with indices around -1.5 suggesting most mass in high-mass clouds

Such a environmental dependence is known from other galaxies eg, Gratier+ found this in M33 or Colombo+ in M51

However, we need to remind ourselves that these cloud samples contain only 50% of the total emission so that massive GMCs cannot contain most molecular material.8

Surface Density of Molecular Clouds2 mass vs. radius for Milky Way and nearby SF galaxy cloudsspiralsdwarfsAndromeda10-kpc ring5-kpc ringstarburstsnot shownouter disk

Surface DensityLimitWith this caveat in mind, I want to show you nevertheless the well-studied scaling relations that Adam already showed you

The first one is the scaling between cloud luminosity and sizeTo put M31 into context, I am showing you the same data for spirals (mostly M51 from PAWS), LGDwarfs (mostly LMC), and outer MW disk (Heyer01)

Over three orders of magnitude you can see a clear scaling with clouds following the dashed lines for fixed surface brighness

However, there is significant scatter indicating that clouds havea range of surface brightness. This scatter exceeds typical measurement uncertainties, therefore Im gonna search for parameters that may explain why clouds are laying here & there

9

Turbulence of Molecular Clouds2 line width vs. radius for Milky Way and nearby SF galaxy cloudsAndromeda10-kpc ring5-kpc ring

Virial RatioGravitational StabilityMaybe the most crucial of Larsons relations is the relationship between velocity dispersion and size of clouds

Under the assumption of virial theorem / energy equipartition, vel dispersion should scale with square root of size (dotted lines) with normalization linked to the clouds mean surface density.

We already saw that the surface density of clouds in not constant, here Im showing you that most clouds are not in virial equilibrium10

Turbulence of Molecular Clouds2 how much of the line width can be explained by self-gravity?Andromeda10-kpc ring5-kpc ringfixed external pressure(Field et al. 2011) concordance of properties (ie offsets) of neighboring clouds?Black lines show scaling expected at fixed external pressure-- right part dominated by internal pressure-- left part dominated by external pressureBut a good fraction of clouds below the zero pressure line thus they are not bound ?!11Concordance of Cloud Properties?3Text

500pc

avg cloud properties (2/l) for neighboring clouds12

DistanceH2 Surface Density @ 200pcCloud Mass3rd7th11th15thSFR Surface Density @ 200pc

Concordance of Cloud Internal Properties?3

Concordance of Cloud-to-Cloud Motions?314

DistanceH2 Surface Density @ 200pcSFR Surface Density @ 200pcCloud Mass

Concordance of Cloud-to-Cloud Motions?3Simulations by Wada et al15

Neighboring Clouds have uncorrelated properties3Molecular Clouds dont have uniform properties2Evidence of diffuse molecular gas in thick disk1A Short Summary16