Haojing Yan

aut viam inveniam aut faciam

Galaxy Stellar Mass Assembly History

At high redshifts

IMHO, one of the most important discoveries of Spitzer was that IRAC detected galaxies at z ~ 6 (and later pushed to a bit higher redshifts). Prior to 2004, nobody would even dare to think that this 85-cm telescope would see galaxies in such an early epoch of the universe (~ 1 Gyr old). In fact, the consensus back then was that one would be lucky if Spitzer could see LBGs at z ~ 3.

It was therefore shocking to see that a z = 5.83 galaxy showed up in the first batch of IRAC Ch1/2 data (3.6/4.5 μm) obtained for the GOODS program. This could only mean one thing (in addition to the happy conclusion that IRAC was so sensitive): the galaxy must be dominated by a matured stellar population that is of rather high mass (several 1E10 solar masses) so that it could emit strongly enough in the rest-frame optical for IRAC to see it in Ch1/2 (sampling at around the restframe V/R). And soon we found two more z ~ 6 galaxies detected in the IRAC data. This implies that the universe must have started forming stars much earlier than z ~ 6 so that at z ~ 6 we could see galaxies of such a high stellar mass! While it was shown a few years ago that some z ~ 3 LBGs have several 1E10 solar masses and are older than 1 Gyr, finding comparable galaxies at z ~ 6 pushed the formation redshift of first galaxies to much earlier.

In Yan et al. (2005, ApJ, 634, 109), we presented our analysis of three z ~ 6 galaxies and eleven z ~ 5 galaxies that were detected in the GOODS IRAC data within the small HUDF. The image stamps and the SED fitting results for the three z ~ 6 galaxies are shown below.

Yan et al. (2005, ApJ, 634, 109)

We soon extended to the whole GOODS fields (north and south). In Yan et al. (2006, ApJ, 651, 24), we presented the analysis on a much large sample, which contains 54 i-dropouts (z ~ 6 candidate galaxies) that are detected in IRAC Ch1/2. Our earlier conclusion was reinforced. The large sample also allowed us to derive the global star formation rate density (GSFRD) at z ~ 6, for which we obtained 1.1-6.7×10⁶ M_solar Mpc^-3 (strictly speaking, this is only a lower limit as we did not correct for the incompleteness of our sample). We also explictly pointed out that the progenitors of such z ~ 6 high-mass galaxies should be quite luminous at higher redshifts (because they must have high SFR), and that they could be rather bright if there was no significant dust extinction.

At intermediate redshifts

We also found some exotic objects in the GOODS IRAC data: they are prominent detections in IRAC, and yet are very weak (or even invisible) in the HUDF ACS (optical) and NICMOS (near-IR) images. These were named "IRAC-selected Extremely Red Objects", or "IEROs" for short. In Yan et al. (2004, ApJ, 616, 63), we discussed 17 such objects. We showed that these red galaxies have the median redshift of z ~ 2.4, and that they dominated by very old stellar populations, with some being as old as the universe at the observed redshifts. In addition, they have surprisingly high stellar masses to the order of 10¹¹M_solar . This is comparable to the highest-mass elliptical galaxies in the local universe. This has two implications: (1) their progenitors could be the earliest galaxies that the universe created, and (2) their progenitors could have gone through a very intense, prolonged period of star formation in order to acquire such high stellar masses.

Yan et al. (2004, ApJ, 616, 63)