Detecting molecules in high-z mergers with ALMA

F. Combes, Observatoire de Paris


Formation and evolution of galaxies

1. Starbursts: increase as a fonction of z

factor 10 between z=0 et 1

2. Dust and extinction:

the most active starbursts are the most obscured (at z=0)

-> at high z, more affected

Star formation rate highly underestimated
LIR/LB varies from 0.1 to 1000 !

3. Negative K Correction

SED for strong and moderate starbursts, at various redshifts, compared with NGST, FIRST/Herschel and ALMA sensitivities, for 1h integration time (Melchior et al. 2001).


JCMT-SCUBA, IRAM-MAMBO at 0.850 -1.2 mm
(Smail et al 97, Barger et al 98, Carilli et al 00)

1-2 sources/ arcmin2
above 1 mJy (e.g. Carilli et al. 2001)

Source number counts, for sub-mm sources, from Carilli et al (2001, astro-ph/0009298)



Identification of sources
Confusion (also SIRTF, ASTRO-F, Herschel)
Contribution of AGN?
How to get redshifts?

By molecular lines (redshift machine)

Extinction corrections applied to vis/UV results correspond ( x 3)

From Genzel & Cesarsky, 2000
Blue losange = ISO
Black squares: without reddening
Green squares: After reddening correction, by a factor 3
Red error bars: SCUBA results
Blue dashed curve: no ULIRG
Red dashed curve: with ULIRG
Black full line: CIBR limits, from COBE

No confusion with ALMA:
spatial resolution better than 0.1"
Possible detection of non ULIRGs
for ex. LBGs: Lyman-Break Galaxies
(Steidel et al 1996, Adelsberger & Steidel 2000)
their density is 150/arcmin2 for z=2.5-3.5

100 times more objects than today

Colour-magnitude diagram (from dust emission and 1600 Angstrom luminosity), for starbursts at z=3, from Adelberger & Steidel (2000, ApJ 544, 218). The x-axis can be viewed as the star formation rate, and the y-axis as the obscuration rate. The blue dots are Lyman-break galaxies, the red and green points are detected sub-mm sources. The black crosses are starbursts at z=0, for comparison. The limit of detection for optical sources is the blue full line, for sub-mm sources is the dash-red line, and for radio sources the dot-green line. The yellow region indicates where very obscured sources can only be detected by their sub-mm or radio emission. With ALMA, the dash-red line will shift to the left by 2 orders of magnitude, and all Lyman-break galaxies will be detected.

Counts compared to theory:

Semi-analytic models, based on the hierarchical scenario
Numerous free parameters

Much more efficient at high z
shorter tdyn?

How to know the molecular gas fraction ?

Molecular lines: favored also at high z
but no negative K correction

Highly depend on excitation, density, temperature..
ULIRGs: observed in excited CO lines
J=8-7, 9-8, etc..

PANORAMA in mm & sub-mm

Tel Area min wavelength Resol
IRAM-30m 707 m2 1mm 10"
IRAM-PdB 883 --> 1060m2 1mm 0.5"
NRO 6x10m=509m2 1mm 0.5"
OVRO 6x10m= 509 m2 1mm 0.5"
+BIMA 10x6m=282 m2 1mm 0.5"
=CARMA 791 m2 1mm 0.5"
SMA 7x6m = 200 m2 0.3mm 0.1"
GBT 100m = 7854 m2 2.6mm 7"
LMT 50m = 1963 m2 1mm 6"
ALMA 64x12=7238 m2 3-0.3mm 0.1-0.01"
E-VLA 35x25m=17200m2 6mm 0.004"



First detection: Faint IRAS Source
F10214+4724 at z=2.3
(Brown & van den Bout 1992, Solomon et al 1992)

CO data for high redshift objects

H2 mass as a function of redshift. Black triangles: normal galaxies. Empty pentagons: ULIRGs. Asterisks: high-z CO-detected objects. Dotted lines: IRAM sensitivity for various CO lines.

Triggered the search: hyper-luminous objects, quasars..

15 betwen z= 1.0 and 4.7 (in 2001)
most of them amplified gravitationally



Starburst modelisation, from z=0 ULIRGs
size 1kpc, mass 6 1010 Mo

2 extreme models:

Assuming the same energy comes from stars

Transfer: LVG model

Result of the LVG model, for the homogeneous case at T=50K. Top: CO lines; Bottom: continuum as a function of redshift.

Starbursts at high z could be different from ULIRGs at z=0

In particular, the molecular density could be less, and the rotational levels of CO not excited until J=9-10..

T=50K   T=30K & 90K
T=30K   HR10, Papadopoulos & Ivison 2001

Molecular absorption in front of quasars will be also a useful tool to determine the chemistry as a function of z, with ALMA

Several examples of molecular absorption in front of quasars, at various redshift (Wiklind & Combes)


Melchior & Combes (2001)

Chemo-dynamical simulations of galaxy collisions at high z

N-body Tree-SPH
with spectral modelisation coupled to dynamics

One of the first simulations, with in green the dark matter, in red the stars, and in blue the gas particules. Three projections are given at each epoch.
In the frame of large-scale cosmological simulations
N-body (Colombi et al)
+ semi-analytical (Guiderdoni et al)

Modelisation of the SED: three types of dust particules are considered
PAH, small and big grains
A population synthesis model is used from the stars formed in the simulation, their light is absorbed by the gas in the same line of sight, within any resolution eleement, and is re-radiated by the dust (from Melchior et al. 2001).

Library of merging galaxies:
the missing link, baryons
star-formation and feedback
disks and angular momentum
dynamical friction..

Observations of the objects:
toward predictions for NGST, ALMA, FIRST-HERSCHEL
and before, GBT, LMT..

Continuum and lines

"Observation" of the simulated galaxies at different redshifts.
Left with NGST
Right with ALMA


Opening of the ALMA "window"