by George Rhee
The distant red galaxies described above may teach us about the origin of the elliptical galaxies we see around us today. Galaxies discovered in the redshift range 1.4 < z < 2 formed most of their ****stars about 1 billion years earlier at redshifts of about three. These objects then merge to form the massive elliptical galaxies that we see today.
What Do High Redshift Galaxies Look Like?
Figure 9.9 shows a set of images of distant galaxies seen in the Hubble Ultra Deep Field. These objects look quite different to any galaxy seen nearby. The first impression is that the high redshift galaxies are more lumpy than nearby galaxies. Scientists must always examine initial impressions critically. Winnie-the-Pooh, who wasn’t a scientist, thought that he had discovered in the snow the footprints of a Heffalump when the footprints were really his own. These Heffalump detections also occur in astronomy.
Fig. 9.9The Hubble Ultra Deep Field gives a glimpse of the morphologies of galaxies at high redshifts. These morphologies are clumpy and irregular (top three rows) in many cases in contrast to spiral and elliptical galaxies. We see top row to bottom row: chain morphologies, clump-cluster, double, tadpole, spiral and finally elliptical galaxies (Credit: Debra Elmegreen et al. 2005, Astrophysical Journal, 631, 85, Reproduced by permission of the American Astronomical Society)
When we use a filter of given wavelength to image a galaxy, the wavelength at which the light that we observe was emitted gets shorter and shorter for high redshifts. We must be careful to compare images of high and low redshift galaxies at similar emitted wavelengths. At high enough redshifts an infrared image of a distant galaxy must be compared with an ultraviolet image of a nearby galaxy. Secondly, the distant galaxy images will only reveal the brightest features that we would see in nearby galaxies. Even when these observing biases are taken into account the fact remains that high redshift galaxies look very different to their low redshift counterparts.
High redshift galaxies have been visually classified into categories such as tadpole, double, chain, chain cluster. None of these morphologies are seen in low redshift galaxies. The clumps seen in high redshift galaxies have masses of 100 million solar masses in stars and sizes of 3,000 light years or less.
We are not quite sure what to make of these unusual galaxy shapes. It is possible that we are witnessing the formation of galactic disks by mergers of smaller halos. Maybe we are witnessing the rapid cosmological infall of gas into a halo. The images suggest that the clumps are regions of intense star formation in disk galaxies.
Cosmic Accounting: Star Formation and the History of the Universe
Can the number of stars we see locally in our universe can be accounted for by counting the number of stars formed since the Big Bang? The lifetimes of stars are much longer than those of human beings. A star having the mass of our sun is expected to shine at roughly constant brightness for about 10 billion years. We thus have a consistency check. We expect the birth rate of stars, that is the star formation rate added up until a certain time to be equal to the density of stars seen at that time.
How then are we to measure the star formation rate of the universe? We measure this as a density since we can’t observe the entire volume of the universe. We take a very large volume of space and use our astronomical observations to estimate how many stars are forming in that volume per year. We then do the same exercise at various redshifts going further back in time. There are various methods used to estimate star formation in galaxies:
The first method is to measure the ultraviolet light emitted by galaxies. Most of that light is emitted by stars having more than five times the mass of the sun. The stars that emit most of the ultraviolet flux only live a few million years so they can’t have been around very long. Their mere presence is an indicator that stars were forming in the ‘recent’ past. We do have to make assumptions about the removal of ultraviolet light by dust to derive the star formation rate from these measurements.
Secondly, we know that recently formed massive stars are hot enough to ionize gas clouds which will emit light. We see the resulting emission lines in the spectra of galaxies, particularly the elements oxygen and hydrogen.
Thirdly, far infrared emission is caused by dust heated by the light of young stars; we can use this infrared flux as a measure of the star formation rate.
The fourth method uses the radio wavelength emission from galaxies to estimate the amount of recent star formation. A 10 solar mass star has a lifetime of 10 million years, one 10,000th of the lifetime of the Sun. These massive stars end their lives in tremendous explosions known as supernovae. What remains after such an explosion is an expanding gas cloud known as a supernova remnant. These supernova remnants are known sources of radio emission, there are some spectacular ones in our own galaxy. In practice we can use the brightness of a galaxy at 20 cm wavelengths to estimate the star formation rate from the supernova rate.
If these different measurements aret all caused by the same phenomenon (star formation) they should be correlated with each other; galaxies that are brighter at 20 cm radio wavelengths should also be brighter in the far infra-red. This is indeed the case.
We can use these measures of star formation rate to estimate star formation history from redshift zero (today) all the way back to high redshifts (Fig. 9.10). In doing this we are plotting the star formation rate over 90 % of the age of the universe. It turns out the star formation rate was much higher 10 billion years ago than it is today. We can also measure the number of stars per unit volume (the stellar mass density) out to redshift four. The stellar density is measured by the near infrared luminosity of these galaxies. It turns out that the stellar mass density increases by a factor 20 from redshift 4 down to redshift 0.
Fig. 9.10 Left panel: the star formation density history of the universe. These measurements are derived from mid-infrared luminosities and sub-mm luminosities of galaxies. Right panel: the stellar mass density of the universe as a function of time and redshift. The star formation rate observed in the left panel is greater by a factor five from what one would infer from the stellar density measurements
Is this increase in the number of stars in our universe consistent with the rate at which stars have been forming over the past 12 billion years? The answer appears to be yes at low redshifts. The addition of all the stars formed in the past 12 billion years accounts rather well for the density of stars we see in the universe today. Half of the mass of stars we see today had already formed by redshift of two, when the universe was 3 billion years old. This is consistent with measurements of star ages in our galaxy. The typical stars in our Milky Way disk were formed 3–6 billion years ago, whereas the oldest stars in our galaxy were formed at least 10 billion years ago.
The density of stars today is computed by adding up the near-infrared luminosity of all the galaxies we see in a nearby volume of space. We choose to use the near-infrared as a tracer of mass in stars. Infrared light is not subject to obscuration by dust. We have to correct for the fact that some of the mass that was formed into stars has returned to the space between the stars. There are two main causes for this, supernova explosions and stellar winds which can cause stars to loose their outer layers. These effects are significant; 30% the mass can be lost in this way. The infrared measurements reveal that about 7% of the baryons in the universe are in the form of stars and the rest are mostly in the form of ionized gas.
At redshifts larger than five it becomes more difficult to estimate the star formation rate and the density of stars and conclusions about consistency and the effects of dust absorption are less reliable.
Cosmic Accounting: The End of Reionization
The reionization of the hydrogen gas in the universe marks the end of the cosmic dark ages and the beginning of the age of galaxies. This is one of the key transitions in the history of the universe yet we know very little about this period. We do know from the spectra of quasars that the universe was reionized by redshift 6. Observations of the cosmic background radiation suggest the
process started at redshift 15 when the universe about 300 million years old.
Where are the sources that reionized the universe? To ionize a hydrogen atom takes an ultraviolet photon. After it is ionized the hydrogen atom will recombine with the electron to form a neutral atom. For a gas of known density and temperature we can calculate the rate at which this will happen. To keep the universe ionized the production of ultraviolet photons capable of ionizing atoms has to match or exceed the rate at which protons and electrons can reform into atoms.
At redshifts less than six the ultraviolet radiation from quasars is sufficient to keep the universe ionized. However at redshifts greater than six, there are not enough quasars around to provide the amount of ultraviolet light needed to ionize the universe. What then is keeping the universe ionized at redshifts higher than six?
Recent observations of galaxies at redshifts of seven suggest that galaxies rather than quasars may have provided the ultraviolet radiation that kept the universe ionized. This then is our second accounting problem. Can we detect the objects responsible for ionizing the universe?
For a star of given mass we know how much ultraviolet light is produced during the lifetime of the star. One can then estimate how many of these stars are required to ionize a given volume of space. Since stars are located within galaxies, wet also need to know what fraction of the ultraviolet light can escape the galaxy and ionize the hydrogen atoms between the galaxies.
By measuring the ultraviolet flux from galaxies we can estimate the rate of star formation up to redshifts as high as seven (Fig. 9.10). Wet also know from Hubble Space Telescope infrared observations how luminous these galaxies are. It seems that most of the ultraviolet light at high redshifts is produced by a large number of very faint galaxies. To find out whether these galaxies produce enough light to keep the hydrogen in the universe ionized new observations of faint distant galaxies are required. We need to image galaxies that are two to three times fainter than the galaxies seen in the longest exposure taken with the Hubble Space Telescope.
The ultraviolet light emitted by electrons going from level 2 to level 1 in hydrogen atoms (known as Lyman-alpha) may yield clues to the reionization question. Observations of high redshift Lyman-α emission from galaxies may help establish when reionization began. At higher redshifts, the fraction of galaxies with detectable Lyman-α emission increases. This suggests that at higher redshifts galaxies have lower dust content as one would expect of the first galaxies. When we go far enough back in time that the universe turns neutral, the Lyman-α emission should be absorbed and this would document the beginning of reionization.
Review
We have presented several techniques that are used to find the most distant galaxies known to us. We have not yet found the first galaxies but we are getting close. We can use the known high redshift galaxies to estimate how the star formation rate changes with redshift. We can also estimate the density of stars at various redshifts. It is instructive to do a consistency check using these quantities, since the sum of the star formation rate over the history of the universe can be used to independently compute the density of stars at any given time. The results seem consistent when dust absorption is taken into account. We have not yet detected the sources of ultraviolet light that are keeping the universe ionized above a redshift of six. A number of puzzles remain, but we are beginning to understand how galaxies were assembled over the history of the universe.
Further Reading
High-Redshift Galaxies. I. Appenzeller. Springer, 2009.
From First Light to Reionization. M. Stiavelli, Wiley-VCH, 2009.
First Light in the Universe. A. Loeb, A Ferrara and R.S. Ellis, Springer, 2008
Observational Cosmology. Stephen Serjeant, Cambridge, Cambridge University Press, 2010
Observing the First Galaxies. James Dunlop, in The First Galaxies - Theoretical Predictions and Observational Clues, Springer, eds. V. Bromm, B. Mobasher, T. Wiklind 2012
George RheeAstronomers' UniverseCosmic Dawn2013The Search for the First Stars and Galaxies10.1007/978-1-4614-7813-3_10© Springer Science+Business Media, LLC 2013
10. Cosmic Archaeology
George Rhee1
(1)Department of Physics & Astronomy, University of Nevada, Las Vegas, Nevada, USA
Abstract
The ancient light that we gather with the largest ground based telescopes together with Hubble Space Telescope data reveals the appearance of galaxies soon after their formation. These galaxies are so distant that we cannot see them in as much detail as nearby galaxies. We can also learn about the distant past by studying nearby galaxies. We can study the age and chemical composition of their stars and use these to reconstruct the history of these systems. In this sense, nearby galaxies are like fossils. In this chapter we present evidence about galaxy formation that we obtain from the study of galaxies very close to us. By close, we mean distances of a few million light years as opposed to the billions of light years that separate us from the galaxies discussed in the previous chapter.
At first I could see nothing, the hot air escaping from the chamber causing the candle flame to flicker, but presently, as my eyes grew accustomed to the light, details of the room within emerged slowly from the mist, strange animals, statues, and gold - everywhere the glint of gold. For the moment - an eternity it must have seemed to the others standing by - I was struck dumb with amazement, and when Lord Carnarvon, unable to stand the suspense any longer, inquired anxiously, “Can you see anything?” it was all I could do to get out the words, “Yes, wonderful things”.
Howard Carter, The Tomb of Tutankhamen
The ancient light that we gather with the largest ground based telescopes together with Hubble Space Telescope data reveals the appearance of galaxies soon after their formation. These galaxies are so distant that we cannot see them in as much detail as nearby galaxies. We can also learn about the distant past by studying nearby galaxies. We can study the age and chemical composition of their stars and use these to reconstruct the history of these systems. In this sense, nearby galaxies are like fossils. In this chapter we present evidence about galaxy formation that we obtain from the study of galaxies very close to us. By close, we mean distances of a few million light years as opposed to the billions of light years that separate us from the galaxies discussed in the previous chapter.
The closest galaxies to the Milky Way lie in a structure known as the Local Group. The local group contains three large spiral galaxies surrounded by numerous dwarf and irregular galaxies. These spiral galaxies are the Milky Way the Andromeda Galaxy and the Triangulum galaxy. We will show evidence of galaxies merging with our Milky Way galaxy and getting ripped apart by tidal forces in the process. We will also show how the populations of stars in nearby dwarf galaxies can be used to measure the formation history of those galaxies. The number of small galaxies surrounding the Milky Way must be explained by our theories. Dwarf galaxies are of interest to us because they represent the extreme faint end of galaxy formation.
The Milky Way Galaxy
The Milky Way has been known for many centuries as a feature of the night sky. The ancient Egyptian and Greeks were aware of it as were native tribes in the Andean foothills and rain forests of south America. Over 400 years ago Galileo used his telescope to reveal that the Milky Way consists of huge numbers of stars. In the late eighteenth century, William Herschel and his sister Carolyn attempted to map the Milky Way by counting stars in different parts of the sky, revealing the Milky Way to be a flattened structure. In 1917 Harlow Shapley calculated distances to a 100 globular star clusters and used these to locate the center of the Milky Way. He concluded that our solar system is located about two thirds of the way from the center of the Milky Way disk. New observing techniques developed in the twentieth century made it possible to observe the Milky Way across the electromagnetic spectrum. At near infrared wavelengths we can see through the dust and see into the disk of the galaxy (Fig. 10.1). Far infrared wavelengths reveal the locat
ion of the dust. Radio observations give a detailed view of the neutral hydrogen gas that is distributed in and above the plane of the galaxy.
Fig. 10.1The entire sky as seen by Two Micron All-Sky Survey. The measured brightnesses of half a billion stars (points) have been combined into colors representing three distinct wavelengths of infrared light: blue at 1.2 μm, green at 1.6 μm, and red at 2.2 μm. This image is centered on the core of our own Milky Way galaxy, toward the constellation of Sagittarius. The two faint smudges seen in the lower right quadrant are our neighboring galaxies, the Small and Large Magellanic Clouds (Credit: Infrared Processing and Analysis Center, Caltech and University of Massachusetts)
The Milky Way is a spiral galaxy with four major components. The first is the dark matter halo whose presence is revealed through its gravity. The total mass of the galaxy is a 1,000 billion (1012) times the mass of the Sun. The second major component is the disk which is about 100,000 light years in diameter and about 3,000 light years thick. The whole disk is embedded in a halo of stars. The disk alone contains 100 billion (1011) stars. It takes 200 million years for our solar system to orbit the center of the galaxy. The fourth component is a bulge in the center of our galaxy that is part of a bar extending from the center.