Low-J CO Line Emission at High Redshift with ALMA Band 2

Leslie HuntINAF-Osservatorio Astrofisico di Arcetri

Firenze, Italy

4mm spectral region rich in lines at z=0

These starbursts, LIRGs (top), ULIRGs (bottom), all show different line ratios (e.g., HCN, HCO+): can we assume the same physical conditions for all these, and at z>0?

From Snell+ (2011), the FCRAO RSR 3mm survey

ALMA fundamental for dusty high-z galaxies

Peak of star-formation activity, Star-Formation Rate Density SFRD, at z ~ 2-3.

LIRGs, Luminous Infra Red Galaxies, 1011 – 1012 L are responsible for >50% of the SFRD from z=0.6 to z=2. ULIRGs account for 20% at z<1.6, increasing to > 50% at z>2.3

(See Reddy+ 2008, Rodighiero+ 2010, Murphy+ 2011).

Dust-corrected UV-derived SFRD from Cucciati+ 2012, VVDS

f gas

z z

Galaxies more gas rich at higher redshifts

Gas fraction fgas = Mgas/(Mgas+M*) depends on XCO (αCO = equivalent mass conversion), the conversion factor from integrated line flux (Tb Δv) to column density NH2.

“Traditional” XCO assumes two values, one for rotationally dominated systems (disks), and another for “starbursts”.

(See Narayanan+ 2012 for continuously varying XCO)

Massive disks vs. starbursts (mergers)

Kennicutt-Schmidt (K-S) relation between SF and gas mass surface densities shows bi-modal behavior or uni-modal behavior depending on CO conversion to total H2 mass.

(Plots taken from Genzel+ 2010; see also Daddi+ 2010, Tacconi+ 2010.)

High-J CO measurements are biased toward warm, dense gas

Spectral-line energy distribution (SLED) of prototypical LIRG starburst, M82 (Weiss+ 2005).

At z> 0.5, typical sub-mm observations trace H2 mass with high-J CO [(3-2), (4-3), …]. These transitions would detect only the inner 400 pc circumnuclear starburst region in M82, rather than the more extended, lower excitation gas.

Uncertainties when CO transitions trace gas mass

Need to assume brightness temperature (Tb) ratios if transitions other than CO(1-0) are used to trace molecular mass.

Need to assume a conversion factor XCO (αCO) to convert CO luminosity to molecular gas mass.

Both parameters depend on physical conditions, including gas volume density n(H2), excitation temperature Tex, dynamical state (e.g., turbulence in clumps vs. rotation dominated), …

Constraint: Using the same XCO for both starbursts and massive disks can sometimes give gas masses which exceed the dynamical mass!

X factor: Relating CO luminosity to H2 mass

CO emission is optically thick (e.g., Wilson+ 1974), hence traces surface area, not volume → need proportionality constant X to relate CO intensity to mass or column density, NH2

Assumptions (e.g., Dickman+ 1986):(1) Extragalactic molecular emission distributed as an ensemble of independent

discrete clouds (no overlap along LOS)(2) Individual clouds virialized (line width ~ dynamical mass)

I(CO) = ∫ Tb dv ≅ Σ TbΔv ~ Σ Tb (M/r)½ ~ Tb Σ (nH2)½ r

~ [Tb (nH2)-½] NH2

Hence, NH2 = X * I(CO), where X ~ (nH2)½ / Tb

virialization mass in homogeneous sphere

NH2 = (nH2) r

Distinct gas phases in z=2 ISM

SLED of lensed Sub-millimeter Galaxy (SMG) at z=2.3. With CO(1-0) from GBT, and CO(3-2), CO(4-3), CO(5-4), CO(6-5), CO(7-6), CO(8-7), CO(9-8) from IRAM 30-m, Danielson+ (2011) find 2 phases necessary: cool, less dense (disk) phase + warm, denser (clumps in starburst) phase.

Factor of 4 temperature variation within various kinematic components!

Jupper

High-J lines underestimate cool gas mass

(Taken from Dannerbauer+ 2009; see also Aravena+ 2010.)

Near-IR selected (BzK) galaxies at z=1.5 mapped in CO(1-0) (VLA) and CO(2-1), CO(3-2) (PdBI) show a SLED in which the cool gas traced by CO(1-0) is underestimated by Jupper > 2. Such SLEDs arise from spatially extended cool, less dense, gas, typical of the low-excitation conditions in quiescent disks (e.g., the local spirals).

Evidence for different K-S relations between different galaxy populations is weak

(Taken from Ivison+ 2011.)

Observational K-S relation with different local and high-z galaxy populations (LIRGs, ULIRGs, BzKs, SMGs) with CO(1-0) (or CO(2-1) measurements.

Offset shown as red arrow for translation when L’CO(1-0) inferred from 3-2 transition.

XCO values inferred from high-J observations could lead to the massive disk vs. starburst dichotomy proposed by several groups. Also lead to steeper slope.

Low-J CO lines key to accurate molecular gas mass: ALMA Band 2 to the rescue

Low-J CO lines key to accurate molecular gas mass: ALMA Band 2 to the rescue

5 transitions with Band 2 from z=0.29 to z=0.72, including CO(1-0).

5-7 transitions with Band 2 from z=1.57 to z=2.44, including CO(2-1).

Band 2 enables low-J SLEDs around the peak of the cosmic SFRD.

ALMA Band 2 will enable accurate gas mass estimates at z>0.3

Band 2 will explore the low-J CO transitions, necessary to accurately infer molecular gas mass and the nature of star-formation activity at high redshift.

This is fundamental for z>0.5 because from z=0, the number of LIRGs increases by almost two orders or magnitude (see Murphy+ 2011).

However, massive star-forming disks with longer gas depletion times, contribute at least 50% of the SFRD at z~ 1.5-2.5 (see, e.g., next slide).

Band 2 will allow comparison of the total molecular gas mass in different galaxy populations, without relying on uncertain Tb ratios or XCO conversion factors.

• Selection of z=0.9-1.9 galaxies from MASSIV (Contini, Epinat, Vergani, … et al. 2011) used to study how disk and spheroidal systems grew through cosmic time (colors show the motions of the Hα gas in the galaxies)

• Need of low-J CO for mass measurements (cold dust and reliable CO/H2 conversion factor) to study the fundamental relations (mass-size-velocity-SFR-metallicity) at low- and intermediate-z

• Need of low-J CO to trace the distribution and kinematics of the (cold, not dense) gas reservoir

Credit: ESO/CFHT; ESO1212 Science ReleaseContini, Epinat, Vergani et al. The Messenger 2012

Thank you!

Last, but not least, 13CO transitions and other potentially optically thin transitions with Band 2 will help resolve degeneracies of temperature/density.