Info

SSP z=7.2 l.Bxlo'Afs 40 Myr : , z=7.8 3.4xlo"lf0 133 Myr " 320 Myr '

i #1417

SSP z=6.7 5.8xlO*Afs 81 Myr ! t100 z= 6.8 9.3xl0"jfs 176 Myr : CSF z=6.B 1.2xlo'°Aia 254 Myr '

SSP z=6.7 5.8xlO*Afs 81 Myr ! t100 z= 6.8 9.3xl0"jfs 176 Myr : CSF z=6.B 1.2xlo'°Aia 254 Myr '

lxlO4 8x10" 3x10" 4xl04 5x10* Wavelength (A)

FIGURE 4.13. Left: Observed SED of the z ~ 7 lensed galaxy from Egami et al. (2005) and model fits from Schaerer & Pello (2005) showing possible solutions with young ages (~15 Myr, solid line) or with a template of a metal-poor galaxy showing strong emission lines. Right: SEDs of two IRAC-detected z ~ 7 galaxies from the Hubble Ultra Deep Field and best fits using three different SF histories. From Labbe et al. (2006). Note the different flux units (Fv versus F\) used in the two plots.

break (Egami et al. 2005; see Figure 4.13). Their analysis suggests that this z ~ 7 galaxy is in the post-starburst stage with an age of at least ^50 Myr, possibly a few hundred million years. If true, this would indicate that a mature stellar population is already in place at such a high redshift. However, the apparent 4000-A break can also be reproduced equally well with a template of a young (~3-5-Myr) burst, where strong rest-frame optical emission lines enhance the 3.6- and 4.5-|im fluxes (Schaerer & Pello 2005, and Figure 4.13). The stellar mass is an order of magnitude smaller (~109M0) than that of a typical LBG, the extinction low, and its SFR ~ M0yr-1.

Two to four of the four z ~ 7 candidates of Bouwens et al. (2004) discussed above have been detected in the very-deep 23.3-h exposures taken with the Spitzer telescope at 3.6 and 4.5 |im by Labbe et al. (2006). Their SED analysis indicates photometric redshifts in the range 6.7-7.4, stellar masses of (1-10) x 109M0, stellar ages of 50-200 Myr, SFRs up to ~25M0 yr and low reddening, AV < 0.4.

Evidence for mature stellar populations at z ~ 6 has also been found by Eyles et al. (2005, 2007). By "mature" or "old" we mean here populations with ages corresponding to a significant fraction of the Hubble time, which is just Gyr at this redshift. By combining HST and Spitzer-telescope data from the GOODS survey they found that 40% of 16 objects for which they had clean photometry exhibit evidence for substantial Balmer/4000 spectral breaks. For these objects, they find ages of ^200-700 Myr, implying formation redshifts of 7 < zf < 18, and large stellar masses in the range ~(1-3) x 101OM0. Inverting the SF histories of these objects leads them to suggest that the past global SFR may have been much higher than that observed for the z ~ 6 epoch, as shown in Figure 4.14. This could support the finding of a relatively high SFRD at z > 7, such as was found by Richard et al. (2006).

In short, although the samples of z > 6 Lyman-break galaxies for which detailed information is available are still very small, several interesting results concerning their properties have emerged already: mature stellar populations in possibly many galaxies indicating redshift (z)

redshift (z)

Figure 4.14. History of the star-formation-rate density determined by inversion from the observed ¿-dropout galaxies analysed by Eyles et al. (2007). The dotted curve is the sum of the past star-formation rates for our ¿'-dropout sample (left axis), with the corresponding starformation-rate density shown on the right axis, corrected for incompleteness including a factor of 3.2 for galaxies below the flux threshold. The dashed curve is this star-formation history smoothed on a timescale of 100 Myr. The triangle is the estimate of the unobscured (rest-frame UV) star-formation-rate density at z ~ 6 from ¿'-dropouts in the HUDF from Bunker et al. (2004). The solid curve shows the condition for reionization from star formation, as a function of time (bottom axis) and redshift (top axis), assuming an escape fraction of unity for the Lyman-continuum photons. From Eyles et al. (2007).

Figure 4.14. History of the star-formation-rate density determined by inversion from the observed ¿-dropout galaxies analysed by Eyles et al. (2007). The dotted curve is the sum of the past star-formation rates for our ¿'-dropout sample (left axis), with the corresponding starformation-rate density shown on the right axis, corrected for incompleteness including a factor of 3.2 for galaxies below the flux threshold. The dashed curve is this star-formation history smoothed on a timescale of 100 Myr. The triangle is the estimate of the unobscured (rest-frame UV) star-formation-rate density at z ~ 6 from ¿'-dropouts in the HUDF from Bunker et al. (2004). The solid curve shows the condition for reionization from star formation, as a function of time (bottom axis) and redshift (top axis), assuming an escape fraction of unity for the Lyman-continuum photons. From Eyles et al. (2007).

a high formation redshift, stellar masses of the order of (1O9-1O1o)M0 and generally low extinction. However, a fraction of these galaxies appears also to be young and less massive (Eyles et al. 2007) forming a different "group". Similar properties and two similar groups are also found among the high-z LAEs (Schaerer & Pello 2005; Lai et al. 2007; Pirzkal et al. 2006) already discussed above. Whether such separate "groups" really exist and, if so, why, remains to be seen.

In a recent analysis Verma et al. (2007) find that >70% of z > 5 LBGs have typical ages of <100 Myr and stellar masses of >109M0, namely are younger and less massive than typical LBGs at z > 3. They also find indications for a relatively low extinction, lower than at z > 3. The trend of a decreasing extinction in LBGs with increasing red-shift has been found in many studies, and is in agreement with the results discussed above for z > 6 and higher. However, the differences in age and mass e.g. compared with the objects of Eyles et al. (2007) may be surprising, especially given the short time (>200 Myr) between redshifts 5 and 6. Several factors, such as selection effects and the representativeness of the small z > 6 samples studied in detail, may contribute to such differences. Reaching a more-complete and coherent understanding of the primeval galaxy types, their evolution and their relation with galaxies at lower redshift will need more time and further observations.

It has been possible during the last decade to push the observational limits out to very high redshift and to identify and study the first samples of galaxies observed barely >1Gyr after the Big Bang. The current limit is approximately at z > 7-10, where just a few galaxies (or galaxy candidates) have been detected, and where spectroscopic confirmation remains extremely challenging.

Thanks to very deep imaging in the near-IR domain it is possible to estimate or constrain the stellar populations (age, SF history, mass, etc.) and dust properties (extinction) of such "primeval" galaxies, providing us with a first glimpse on galaxies in the early Universe. Despite this great progress and these exciting results, the global observational picture on primeval galaxies, and on their formation and evolution, remains to be drawn. Many important questions remain or, better said, are starting to be posed now, and can now or in the near future be addressed not only by theory and modelling but also observationally!

We have already seen some of the emerging questions, but others, sometimes more-general ones, have not been addressed. Among the important questions concerning primeval galaxies we can list the following.

• How do different high-z populations such as LAEs and LBGs fit together? Are there other currently unknown populations? What are the evolutionary links between these populations and galaxies at lower redshift?

• What is the metallicity of the high-z galaxies? Where is Population III?

• What is the star-formation history of the Universe during the first Gyr after the Big Bang?

• Are there dusty galaxies at z > 6? How, where and when do they form? How much dust is produced at high redshift?

• Which are the sources of reionization? Are these currently detectable galaxies or very-faint low-mass objects? What is the history of cosmic reionization?

We, and especially young students, are fortunate to live during a period when theory, computing power and observational facilities are rapidly growing, enabling astronomers to peer even deeper into the Universe. It is probably a fair guess to say that within the next 10-20 years we should have observed the very first galaxies forming in the Universe, found Population III, etc. We will thus have reached the limits of the map in this exploration of the Universe. However, a lot of challenging and interesting work will remain if we are to reach a global and detailed understanding of the formation and evolution of stars and galaxies!

Was this article helpful?

0 0

Post a comment