z=0Modelling the atomic and molecular gas in galaxies and comparisons with GASS and COLD GASS results

With Jian Fu, Qi Guo, Cheng Li, and the COLD GASS and GASS teams

DavidSchiminovich (PI)

Barbara Catinella

Amelie Saintonge

GALAXY SELECTION a) Redshift range : 0.025 < z < 0.05Motivation: 1) We want to detect HI and H2 down to levels of a few percent inM(gas)/M* in about an hour of integration time. 2) We want to get an accurate estimate of the total cold gas mass with a single pointing of both the Arecibo and IRAM telescopes.

b) Stellar mass range: log M* > 10Motivation:1) Span a range of stellar masses encompassing both “active” star-forming galaxies and “passive” systems => quantify how the transition between thesetwo populations is reflected in their cold gas content. Avoid any selection on morphology, environment etc.

2) In the mass range, all galaxies have roughly solar metallicity: conversionfactor from CO luminosity to H2 mass is simplified.

OBSERVING STRATEGY: Integrate until the galaxy is detected or a limitingHI/H2 mass fraction ~1.5-3% is reached, i.e. Our survey will quantify the fullcondensed baryon budget in these galaxies.

MILLENNIUM-II Simulations (Boylan-Kolchin et al 2009)

THE MERGING TREE

Models are based on the L-Galaxies Code developed by Springel, Croton, De Lucia, Guo et al.

Radially Resolved Models for Galactic Disk Formation

Scale radius of the infalling gas is proportional to λ Rvir (following Fall & Efstathioiu 1980; Mo, Mao, White 1998)

The gas that falls is at one snapshot is simply superposed on the gas that has already fallen in at earlier times.

Krumholz, McKee and Tumlinson 2009 Blitz & Rosolowsky 2004,2006

GAS PROFILES ARE TOO FLAT AND THE MOLECULAR GAS IS RAPIDLY CONSUMED IN THE INNER DISK.

PROBLEM CANNOT BE SOLVED BY ADJUSTING FREE PARAMETERS.

Comparison of Model Profiles with THINGS/ HERACLES data from Bigiel et al 2008

IF ENERGY FROM SUPERNOVA IS MORE EFFICIENTLY DISSIPATED IN DENSE INNER REGIONS OF THE DISK (i.e. LESS COLD GAS IS REHEATED ), THEN MODELS CAN FIT DATA.

Caveat: Radial inflow of gas has not been considered

Comparison of Model Profiles with THINGS/ HERACLES data from Bigiel et al 2008

The Models Provide a Reasonable Match to the Observed Mass Functions of HI, H2 and Stars in the Local Universe

It is interesting to consider the global scaling relations between stellar mass, stomic gas mass and molecular gas mass for an ensemble of disk galaxies

What Are the Physical Drivers of these Scaling Relations?1) The MASS OF THE DARK MATTER HALO determines the mass of baryons that is able to cool. Because supernovae feedback is less efficient in massive halos, a larger fraction of the baryons that cool are converted into stars.

What Are the Physical Drivers of these Scaling Relations?

1) The SPIN PARAMETER sets the contraction factor of the infalling gas. A higher spin parameter results in larger galaxies with lower gas surface densities.

What Are the Physical Drivers of these Scaling Relations?

1) The FRACTION OF GAS in the galaxy that was ACCRETED recently (i.e. Gas that has has less time to be turned into stars).

COMPARE WITH DR2 FROM THE

COLD GASS SURVEY

(300 galaxoes with both HI observations from Arecibo and CO 1-0 observations from the IRAM 30m telescope)

BLUE: Arecibo/IRAM detections RED: Arecibo/IRAM non-detections

PROCEDURE:

1) Analyze detections and non-detections separately. 2) Analysis of population of galaxies with detected HI/CO allows us to test the disk formation models . 3) Analysis of galaxies without detected gas allows us totest whether “quenching” processes (in particularradio-mode AGN feedback) operates the same way in themodels and in the data.

Comparison of Mean HI and H2 scaling relations

Atomic Gas Molecular Gas

Saintonge et al 2011

Molecular gas depletion times are shorter in more actively star-forming galaxies

Distribution Functions of the HI detections in bins of mass, stellar mass density, concentration

Trends in non-detected fraction do NOT agree with the models

Fraction of non-detections is independent of stellar mass in the data, but depends strongly on stellar surface density andconcentration (i.e. Bulge-to-disk ratio)

In the models, “radio-mode” AGN feedback prevents gas from cooling in massive galaxies (with black holes) located in halos above ~10^11 M_sun. Observationally, this is manifested as a strong threshold in STELLAR MASS

H2 detections and non-detections in 2-dimensional planes of stellar mass and galaxy structural parameters

Natural interpretation of the results: processes associated with the formation of galactic bulges and

or black holes are most directly responsible for shutting down star formation in galaxies.

PHOTOMETRIC INDICATORS OF ATOMIC GAS CONTENT

UGC 8802: A Case Study of a Star-Forming Outlier from the HI “plane”, with unusually high HI fraction

Moran et al 2010

M(HI) = 2 x 10 10 M_sol

As the HI gas fraction increases, galaxies become bluer on the outside relative to the inside. This effect is weaker in control samples matched in stellar mass, structural parameters and specific star formation rate.

Wang et al 2011

Preliminary work with Cheng Li:Find best photometric “predictors” of the HI content of a galaxy

Update of HI gas fraction “plane” to include a colour gradient term --gives much more accurate predictions for high Hi fraction galaxies.

GASS-SDSS cross-correlation functions: “real” versus “predicted”

What is this useful for? Creating mock catalogues – tuning up future surveys with real data!