by George Rhee
The hope is that we will detect events such as the formation of the first stars (30 > z > 22). The first black holes in the universe may have formed in the redshift range 22 > z > 13 as remnants of the first stars. The effect of these black holes would be to transform the 21-cm signal from absorption into emission. As most of the gas becomes ionized (13 > z > 6) the signal from neutral hydrogen gas is no longer detectable. The spectral information can potentially tell us when the first stars formed in the universe, when the first black holes formed, when reionization began and when it ended.
To image these events we need to use radio interferometers. An interferometer combines the signal of many individual radio receivers. Radio telescopes being built for this purpose include the low frequency array (LOFAR) in the Netherlands which has 44,000 antennas covering an area of 1,500 km diameter (most of the antennas will be in the Netherlands but a few will be placed in France, Germany, the UK and Sweden). The Murchison wide field array located in the radio quiet western Australian outback will consist of 2,048 antennas with separation of up to 3 km.
Figure 8.5 shows what one might expect the redshifted 21-cm line sky to look like. The images cover an area of the sky comparable to the full moon. The redshift range is 18.5 to 11. At first the sky is covered by neutral hydrogen seen in emission (shown in green in the figure). Gradually the neutral hydrogen (green) gets converted to ionized hydrogen (shown in orange). At a redshift of eleven (bottom right panel) there is almost no detectable neutral hydrogen.
Fig. 8.5Slices through a model universe simulating the reionization process. The neutral hydrogen is shown in green, the ionized hydrogen in orange. The redshifts of each slice in the top row from left to right are 18.5, 16.1 and 14.5. The redshifts of each slice in the top bottom from left to right are 13.6, 12.6 and 11.3. This covers a period from 250 million years after the Big Bang to 400 million years after the Big Bang. The angular size of this box viewed on the sky would be roughly one and half times the size of the full moon. By a redshift of 11.3 less than 1 % of the neutral hydrogen remains. Reionization proceeds through the overlap of ionized bubbles (Credit: I. T. Iliev, G. Mellema, U.-L. Pen et al. Simulating cosmic reionization at large scales I. The geometry of reionization, Monthly Notices of the Royal Astronomical Society, (2006), 369, 1625–1638, by permission of Oxford University Press on behalf of the Royal Astronomical Society)
The instrument best suited to carrying out these mapping measurements is the Square Kilometer Array (SKA) to be located in South Africa and Australia. The SKA will search for neutral hydrogen in the redshift range from 20 to 7, precisely where the reionization transition is expected to take place (Fig. 8.4). The total collecting area of the SKA will be 1 km2. To achieve this, the SKA will use 3,000 dish antennas, each about 15 m wide as well as two other types of radio wave receptor, known as aperture array antennas. The antennas will be arranged in five spiral arms extending to distances of at least 3,000 km from the center of the array. The central regions in Australia and in South Africa will contain cores each 5 km in diameter; one for each antenna type. The aperture array antennas will extend to about 200 km from the core regions. In Southern Africa the dishes will be positioned in distant stations out to at least 3,000 km.
The SKA will be a factor of ten more sensitive than the currently most sensitive instrument, the Expanded Very Large Array radio interferometer. The first observations will be carried out in 2019 and the full array will be operational by 2024. New computer systems will be necessary to deal with the vast amount of data produced by SKA. The radio antennas will produce a flow of data of up to 1015 bits per second, which is 100 times the total amount of internet traffic today.
We havet already mentioned that part of the trouble in making low frequency measurements from the ground lies in what is called radio frequency interference from man-made sources such as television transmission and cell phones. The FM band used for broadcasting radio goes from 90 to 1,100 MHz which is precisely where the effects of reionization are strongest. The radio frequency range from 30 to 110 MHz has been almost exclusively allocated to cell phone usage and TV and radio. This makes it necessary to locate the telescopes in parts of the world, such as the deserts of Australia, that are radio quiet.
The study of the dark ages is a field that holds great promise. Astrophysicists are working hard to predict the properties of neutral hydrogen prior to and during the epoch of reionization. We hope to answer several questions. When did reionization take place? How long did it last? What sources were responsible? The hope is that in the next decade we will go from the realm of speculation to anchoring our theoretical ideas on sound physical measurements. At redshifts less than seven we can already observe neutral hydrogen in the universe using the spectra of quasars.
Neutral Hydrogen at z < 7: Evidence from Quasar Spectra
Optical spectra of quasars provide evidence that neutral hydrogen exists at high redshifts. The electron in a hydrogen atom can go from one orbit to another by absorbing and emitting light at certain known wavelengths. When hydrogen gas is cold, the electrons aret all in the lowest orbit. The electron can jump to the next energy level by absorbing ultraviolet light with a wavelength of 1,216Å. If the gas is hot (but not ionized) the gas will emit light at that wavelength. We see this light as a feature in quasar spectra, it is referred to as the Lyman-α line (Figs. 8.6 and 8.7).
Fig. 8.6Quasar spectra tell us about the gas intervening between us and the quasar because hydrogen gas removes light from the quasar spectrum at known wavelengths. These wavelengths get shifted towards the red telling us about the distance from us of the hydrogen gas clouds. The solid red line marks the quasar spectrum at wavelengths longer than the peak of emission. The dashed line shows what we think the spectrum would look like in the absence of hydrogen absorption (Credit: John K. Webb, University of New South Wales)
Fig. 8.7The panel compares the spectra of two quasars at very different redshifts. 3C 273 is at a redshift of 0.158 whereas 1, 422 + 2, 309 is at a redshift of 3.62. The spectra are shown at their emitted wavelengths. The strong peak is the Lyman-α emission line. It is visible in both spectra. In the high redshift quasar spectrum the emission blueward (to the left) of the peak is lessened by the presence of many narrow absorption lines of neutral hydrogen known as the Lyman-α forest (Credit: Michael Rauch, Carnegie Institution of Washington, Sally Heap, Space Telescope Science Insitute, NASA/ESA, Bill Keel, University of Alabama)
Quasars are among the most luminous sources known to us. The light is emitted by gas spiraling into a black hole at the center of a galaxy. What we see as a quasar is the compact region in the center of a massive galaxy surrounding its central black hole. More than 200,000 quasars are now known. These were mostly discovered by the Sloan Digital Sky Survey. The most distant quasars have redshifts of about seven so their light was emitted less than 1 billion years after the Big Bang.
Figure 8.6 illustrates how the spectrum of a quasar reveals the presence of intervening hydrogen gas. Figure 8.7 shows spectra of two quasars at different redshifts. The spectra are shown in their rest frame so that the Lyman-α emission is at the same location on the x-axis.
At wavelengths longer than 1,216 Å the spectra look somewhat similar but at shorter wavelengths they look completely different. The higher redshift quasar (lower panel) shows many sharp dips. These dips are collectively known as the Lyman-α forest. Each dip is a feature known as an absorption line that is believed to be caused by a cloud of neutral hydrogen gas that lies between us and the quasar. Comparing these two spectra tells us that in the past there were many more such clouds than in more recent times. By comparing spectra of quasars at different redshifts we can probe the distribution of neutral hydrogen at various epochs during the last 13 billion years. As we go to redshifts larger than six something dramatic happens. Figure 8.8 illustrates what the intrinsic spectrum of the quasar would look like if there were no intervening neutral hydrogen and what it ac
tually looks like. The abundance of neutral hydrogen is high enough that the light blueward of Lyman-α has almost entirely been absorbed.
Fig. 8.8Comparison of the observed spectrum of a quasar at redshift 6.37 (black line) with the expected spectrum based on low redshift quasar observations (dashed red line). To the left of the Lyman-α peak there is essentially no emission suggesting that there is sufficient neutral hydrogen present to entirely absorb the light from the quasar. Note however that it does not take much neutral hydrogen to produce this effect (Credit: Richard White, Robert Becker, Xiahui Fan and Michael Strauss, 2003, Astrophysical Journal, 126, 1, reproduced by permission of the American Astronomical Society)
Figure 8.9 shows the spectra of 18 quasars in the redshift range 5.7 < z < 6.4. The quasar spectra show that the neutral hydrogen fraction in the universe is changing with time. It is changing in the way that we expect with more neutral hydrogen present in the distant past. The data support the idea that the epoch of reionization is over by redshift six. It is not clear at present what objects were responsible for reionizing the universe. There are not enough quasars at redshift six to do the job. It is possible that star forming galaxies are responsible since young massive stars producet alot of ultraviolet radiation. The galaxies that we have detected at high redshifts aret also not sufficient to do the job. Maybe an as yet undetected population of high redshift dwarf galaxies could do it. This is plausible because in the dark matter picture of galaxy formation the smaller galaxy halos form first.
Fig. 8.9The spectra of 18 quasars discovered with the Sloan Digital Sky Survey. The spectra are ranked from bottom to top in order of increasing redshift from a redshift of 5.74 for the bottom spectrum to a redshift of 6.42 for the top spectrum. The wavelength is plotted on the horizontal axis and the intensity on the vertical axis. As we go to higher redshifts there is less and less light remaining to blueward (to the left) of the Lyman-α emission line. In the quasars near the top of the figuret almost all the light has been absorbed by neutral hydrogen gas located between us and the quasar. What we are seeing is a transition from an opaque universe (redshifts greater than six) to a transparent universe at lower redshifts, reflecting the end of the dark ages. For comparison see the quasar spectra in Fig. 8.7 at much lower redshifts of 0.16 and 3.6 (Credit: Fan et al. 2006, Astronomical Journal, 132, 117, reproduced by permission of the American Astronomical Society)
Review
It may not be possible for us to directly detect the very first stars and galaxies to form in the universe. Our knowledge of their formation will then depend on the effect that they have on their surroundings. In that case, mapping and spectroscopy of neutral hydrogen at high redshift will be the only way to infer that the first generation of stars are forming. We have learned how the first stars will modify the spectrum of the cosmic background radiation that we measure at meter wavelengths. We have seen what an image of the neutral hydrogen sky might look like at the various wavelengths covered by the epoch of reionization. We learned how the quasars provide beams of light that can be used to map the neutral hydrogen clouds that survived the reionization of the universe. We measure the size of these clouds, how much neutral hydrogen they contain and their abundance of elements other than hydrogen. We can also study the clustering of these clouds which contains information on the dark matter distribution in the universe.
It would be most exciting if we could directly detect the earliest galaxies. We turn in the next chapter to the story of our attempts to push back to earlier and earlier times our observations of stars, galaxies and black holes.
Further Reading
The Reionization of Cosmic Hydrogen by the First Galaxies. Abraham Loeb in Adventures in Cosmology. David Goodstein, ed., Singapore, World Scientific. 2012.
Reionizing the Universe with the First Sources of Light. Steven Furlanetto in Adventures in Cosmology. David Goodstein, ed., Singapore, World Scientific. 2012.
How did the First Stars and Galaxies Form? Abraham Loeb. Princeton. Princeton University Press. 2010.
Observational Cosmology (Chapter 8, The Intervening Universe). Stephen Serjeant. Cambridge. Cambridge University Press. 2010.
George RheeAstronomers' UniverseCosmic Dawn2013The Search for the First Stars and Galaxies10.1007/978-1-4614-7813-3_9© Springer Science+Business Media, LLC 2013
9. Observing the First Galaxies
George Rhee1
(1)Department of Physics & Astronomy, University of Nevada, Las Vegas, Nevada, USA
Abstract
Astronomers today are close to seeing galaxies as they appeared just after the formation of the first stars. In this chapter we present three techniques currently used to find the most distant galaxies. The first two reveal galaxies that are forming stars at several hundred times the rate of the Milky Way galaxy. The third method selects galaxies with stars older than a billion years. We then discuss two accounting problems. We can measure the star formation rate over a large part of the history of the universe, but can we reconcile this birth rate with the number of star we actually see today? Secondly we know that sources of ultraviolet radiation kept the universe ionized at redshifts larger than six. We know that at lower redshifts the radiation from quasars is sufficient to do the job. At redshifts larger than six we do not see enough quasars to do the job but could galaxies be responsible?
With increasing distance our knowledge fades, and fades rapidly. Eventually we reach the dim boundary - the utmost limits of our telescopes. There, we measure shadows, and we search among ghostly errors of measurement for landmarks that are scarcely more substantial.
Edwin Hubble
Astronomers today are close to seeing galaxies as they appeared just after the formation of the first stars. In this chapter we present three techniques currently used to find the most distant galaxies. The first two reveal galaxies that are forming stars at several hundred times the rate of the Milky Way galaxy. The third method selects galaxies with stars older than a billion years. We then discuss two accounting problems. We can measure the star formation rate over a large part of the history of the universe, but can we reconcile this birth rate with the number of star we actually see today? Secondly we know that sources of ultraviolet radiation kept the universe ionized at redshifts larger than six. We know that at lower redshifts the radiation from quasars is sufficient to do the job. At redshifts larger than six we do not see enough quasars to do the job but could galaxies be responsible?
Astronomy and Geology
We are getting to the point where we can observe the end of the epoch of reionization described in the last chapter. About 30 years ago our telescopes could study galaxies out to a redshift of one. We see these galaxies as they appeared about 6 billion years after the Big Bang. After the launch of the Hubble Space Telescope, it was realized that by making long exposure images of the sky one could see much further. The famous image known as the Hubble Deep Field pushed the redshifts at which we could image galaxies to about four. We see these galaxies as they were 1.5 billion years after the Big Bang. In 2004, the Hubble Ultra Deep Field made it possible to look back over 13 billion years of the history of the universe to see galaxies as the appeared only 700 million years after the Big Bang. This involved pointing the Hubble Space Telescope for several days at an empty part of the sky about one tenth as large as the full moon. The infrared Hubble Ultra Deep Field pushed to about 600 million years after the Big Bang and maybe fainter to redshifts as large as 10. This is the state of the art. Hubble’s successor, the James Webb telescope which we shall discuss at length in Chap. 12 is expected to probe to redshifts of 20, less than 200 million years after the Big Bang (Fig. 9.1). Two hundred million years is a very long time in our everyday world but it is a small fraction (1.5 %) of the age of the universe. By exploring this new territory wet also expect to discover entirely new phenomena. This has been the pattern throughout the history of astronomy.
Fig. 9.1As we probe the distant universe we observe galaxies with increasingly large redsh
ifts. Each redshift value corresponds to a time after the Big Bang at which the light was emitted as shown by the upper axis labels in the figure. The most distant galaxies are seen in very long exposures taken with cameras on the Hubble Space Telescope. The Hubble Space Telescope imaged a very small region of the sky repeatedly for 35 h in 1995. The the experiment was repeated with a new camera on Hubble, with 55 h exposures in 2004. The 1995 experiment is called the Hubble Deep Field. The 2004 measurements are known as the Hubble Ultra Deep Field (Fig. 9.6). In 2010 an infrared version of the Hubble Ultra Deep Field imaged galaxies that may have redshifts as high as ten. The goal of the James Webb Telescope is to extend this work out to redshifts greater than 20. Each redshift value corresponds to a time after the Big Bang at which the light was emitted as shown by the upper axis labels in the figure (Credit: Rychard Bouwens)
The science of geology is similar to observational cosmology. Geologists study the sedimentary layers of the Earth’s rock to tell the story of the Earth across hundreds of millions of years of its history. In the southwestern United States the rocks tell a remarkable story of river deltas and the ebb and flow of vast inland seas in stark contrast to the desert landscapes which we see around us today. In a similar manner our telescopes show us that galaxies in the distant past were quite different from the ones we see around us today.
Cosmology is based on the assumption that we live in an ordinary region of the universe and that, averaged over large enough volumes, the universe is essentially the same everywhere. If this is the case then observations of very faint distant galaxies seen at very early times give us a picture of what our own galaxy and its neighbors would have looked like at comparably early times. The challenge in studying these very distant galaxies lies in the fact that they are extremely faint and that their visible light is shifted into the infrared by the time it reaches our telescopes.