Judd D. Bowman
Assistant Professor
School of Earth and Space Exploration (SESE)
Arizona State University (ASU)
Return to homepage

Epoch of Reionization

Despite the amazing progress of observational astrophysics and cosmology over the last decade, a large gap remains in the history of the Universe for which there are currently no direct probes. The period between 370,000 and 1 billion years after the Big Bang, referred to as the cosmological Dark Ages, is completely invisible to contemporary observations. Information from this period is not explicitly contained in the cosmic microwave background (CMB) because baryonic matter and radiation have already decoupled and CMB photons stream freely through the intergalactic medium (IGM) during this epoch. Yet, there has been insuficient time for significant numbers of stars, galaxies, and quasars to form and produce emission that could be detected with the telescopes available today. Not coincidentally (and almost by definition), the existing capabilities to detect high-redshift galaxies and quasars reach only to the very end of the Dark Ages.

Probing this epoch is at the forefront of modern astrophysics and cosmology. It is a daunting challenge to study the sources of the first light, but several options do exist. According to theory, the first generation of stars could form as early as 100 to 200 million years after the Big Bang. These stars will be extremely massive, of order 100 solar masses, and therefore very short-lived. In principle, their deaths should produce prodigious numbers of gamma-ray bursts that may eventually be detected. Additionally, extremely over-dense regions in the Universe will collapse particularly early due to gravitational instability, and thus rare, but massive galaxies and quasars may exist, even during early times. The James Webb Space Telescope (JWST), scheduled for launch in 2013, and next-generation 30 m ground-based telescopes should be able to detect these objects out to very high redshifts (z < 15). Finally, there is a third major potential probe of the Dark Ages. The bulk of the baryonic matter in the Universe during this period is in the form of neutral hydrogen gas in the IGM. Rather than target observations at the galaxies and quasars that are the rare, early products of gravitational collapse, it should be possible to detect directly the presence of the ubiquitous hydrogen gas. The most promising method of achieving this detection is to search for signatures of the (highly redshifted) 21 cm hyperfine transition line of neutral hydrogen in the radio spectrum. These observations will be one of the key science goals of the Square Kilometer Array (SKA) when it is built around 2020, but precursor experiments should be able to make significant progress in the near future.

The transition period at the end of the Dark Ages is known as the epoch of reionization (EOR). During this epoch, radiation from the very first luminous sources--early stars, galaxies, and quasars--succeeded in ionizing the neutral hydrogen gas that had filled the Universe since the recombination event that occurred as the Universe cooled following the Big Bang. Reionization marks a significant shift in the evolution of the Universe. For the first time, gravitationally-collapsed objects exerted substantial feedback on their environments through electromagnetic radiation, initiating processes that have dominated the evolution of the visible baryonic Universe ever since. The epoch of reionization, therefore, can be considered a dividing line, of sorts, when the relatively simple evolution of the early Universe gave way to more complicated and more interconnected processes. This is the period that will be probed most thoroughly by the next generation of optical, infrared, and radio observatories.

Redshifted 21 cm Background

New experiments are aimed at detecting redshifted 21 cm signatures against the CMB. This signal from the Dark Ages appears as a faint, diffuse background in radio frequencies below 200 MHz (for redshifts above z = 6). Measuring the brightness temperature of the redshifted 21 cm background could yield information about both the global and local properties of the IGM. Determining the average brightness temperature over a large solid angle as a function of redshift would eliminate any dependence on local density perturbations and constrain the evolution of the product x_HI (1 - T_CMB/T_S), where x_HI is the global neutral fraction of hydrogen, T_CMB is the CMB temperature, and T_S is the spin temperature of neutral hydrogen in the IGM. During the reionization epoch, it is, in general, a good approximation to assume that the spin temperature of neutral hydrogen is much greater than the CMB temperature (T_S >> T_CMB) and, therefore, that the brightness temperature is proportional directly to x_HI . Global constraints on the brightness temperature of the redshifted 21 cm line during the EOR, therefore would directly constrain the neutral fraction of hydrogen in the IGM. This would yield significant improvements in estimates of the optical depth to CMB photons and, thus, would help to break existing degeneracies in CMB measurements between the optical depth and properties of the primordial matter density power spectrum [Tegmark et al., 2006]. They would also provide a basic foundation for understanding the astrophysics of reionization by setting bounds on the duration of the epoch, as well as identifying unique features in the ionization history (for example if reionization occurred in two phases or all at once).

Results of a simulation illustrating fluctuations in the observable brightness temperature due to redshifted 21 cm emission near the end of the reionization epoch when the ionized fraction of gas is only x_i=0.3. Data courtesy of A. Mesinger & S. Furlanetto, 2007. (166 KB, jpg)

On small scales, local perturbations in the density, ionization fraction, or spin temperature may produce significant deviations from the typical, global averages of the observed brightness temperature. Characterizing these fluctuations would be a powerful approach to exploiting the information in the redshifted 21 cm background. As primordial hydrogen cools following recombination and later reheats, density contrasts in the baryonic matter distribution should be revealed as fluctuations in the brightness temperature of the redshifted 21 cm line [Sunyaev and Zeldovich, 1972, Hogan and Rees, 1979, Scott and Rees, 1990, Iliev et al., 2002, 2003, Loeb and Zaldarriaga, 2004, Barkana and Loeb, 2005b]. At high redshifts prior to reionization, fluctuations in the redshifted 21 cm background are expected to follow closely the matter density fluctuations -- at a time when baryon perturbations were still substantially in the linear regime|and should contain information regarding the fundamental cosmological model. Redshifted 21 cm observations may help constrain the geometry of the high redshift universe between recombination and reionization [Barkana and Loeb, 2005c]. In particular, at very small spatial scales, where neither CMB anisotropy measurements nor large-scale structure surveys are able to directly probe the matter power spectrum, redshifted 21 cm measurements may dramatically improve constraints on alternatives to the standard inflationary model. Loeb and Zaldarriaga [2004] calculate that the number of independent modes accessible through redshifted 21 cm measurements of the matter power spectrum is up to nine orders of magnitude greater than for CMB measurements.

During the reionization epoch (z < 15), a unique pattern will be imprinted in the redshifted 21 cm signal that reflects the processes responsible for the ionizing photons and that evolves with redshift as reionization progresses. As the first luminous sources ionize their surroundings, voids are expected to appear in the fluctuating emission [Madau et al., 1997, Tozzi et al., 2000, Ciardi and Madau, 2003, Zaldarriaga et al., 2004, Furlanetto et al., 2004b]. In principle, these features may be studied through direct imaging or through the determination of the power spectrum (and higher-order statistics) of the spatial fluctuations. The fluctuations may be probed along a single line-of-sight (resulting in spectral features that are similar to the Lya forest and that are dubbed the 21 cm forest), angularly in the plane of the sky (yielding maps of the fluctuations like those produced by WMAP for the CMB), or three-dimensionally in an observed volume of space. The last method gives rise to 21 cm tomography.

The properties of the three-dimensional power spectrum of the spatial fluctuations in the redshifted 21 cm brightness temperature are expected to be dominated by the characteristics of the reionized voids in the background emission due to the first luminous objects [Zaldarriaga et al., 2004, Furlanetto et al., 2004a]. Measurements of the power spectrum from this period would provide insight into many of the poorly understood processes responsible for reionization and structure formation, such as the radiative feedback mechanisms in star-forming regions, the physics of the first (Population III) stars, and the role of quasars. Tracing the power spectrum as a function of redshift during this epoch will chart the history of the formation of structures. Directly imaging voids in redshifted 21 cm brightness temperature from individual HII regions surrounding quasars during this epoch would probe quasar physics [Wyithe and Loeb, 2004a,c, Kohler et al., 2005] and could provide guides in searches for high-redshift galaxies [Wyithe et al., 2005]. Measurements of the redshifted 21 cm background during the EOR may also be useful for constraining cosmological models. Ali et al. [2005] and Barkana and Loeb [2005a] have shown that differences in the line-of-sight versus angular components of the observed redshifted 21 cm power spectrum can be used to separate primordial density perturbations from features caused by the radiative processes responsible for reionization, and Barkana [2006] has discussed the application of the Alcock-Paczynski (AP) test [Alcock and Paczynski, 1979] to redshifted 21 cm measurements. Additionally, Barkana and Loeb [2005b] consider the effects of the earliest galaxies on the redshifted 21 cm fluctuations and Naoz and Barkana [2005] discuss using redshifted 21 cm observations to study the thermal history of hydrogen gas by detecting a small-scale cutoff in the power spectrum due to thermal broadening of the hyperfine line.

Experimental Approaches

There are two broad approaches to using the redshifted 21 cm signal to probe the Dark Ages. The first approach is to constrain the global evolution of the average redshifted 21 cm differential brightness temperature with redshift, and the second is to characterize the local fluctuations in the background. Both types of observations would provide important information about the Dark Ages and the epoch of reionization. In principle, the easier approach to observing the redshifted 21 cm background is to chart the evolution of the global differential brightness temperature with redshift. Since the goal in this case is to average over a large solid angle at multiple frequencies, global signature experiments do not need necessarily to image the sky. Furthermore, since the redshifted 21 cm signal is visible in all directions, the signal will fill the primary beam of any antenna. This provides a significant simplification and means that there is no loss in sensitivity by increasing the field of view (as would be typical if one were observing a point source). Thus, global signature experiments are able, in principle, to use very simple antennas, such as individual dipoles.

Several small experiments are underway that are designed to detect distinct features in the global redshifted 21 cm background [Shaver et al., 1999, Gnedin and Shaver, 2004, Furlanetto, 2006], such as a sharp step transition in the all-sky spectrum that would be present if reionization occurred very rapidly. Amazingly, these modest experiments could have been performed easily (for the most part) anytime in the past few decades, but were not conceived until significant attention was turned to understanding the reionization epoch. Two primary efforts in this category are the Compact Reionization Experiment (CORE) led by Dr. Ron Ekers at the Australian Telescope National Facility, and our Experiment to Detect the Global EOR Signature (EDGES), a collaboration with Dr. Alan Rogers at the MIT Haystack Observatory.

First deployment of the EDGES system at Mileura Station in western Australia in December, 2006. The antenna and groundscreen are seen in the foreground, while the receiver and analog-to-digital conversion unit is visible in the background. The two components are connected by a 50 meter transmission line and several control cables. Alan Rogers stands ready to adjust the configuration. (100 KB, jpg)

The second approach discussed above to observing the redshifted 21 cm background is to characterize the fluctuations in the signal. Unlike the global signature efforts, experiments using this approach are required to provide information about the sky on small angular scales. The ideal outcome of observations for this approach would be true maps of the redshifted 21 cm background. However, due to the extremely intense synchrotron radiation from our own galaxy, directly imaging the fluctuations and voids in the redshifted 21 cm background with the desired arc-minute or better resolution will require the sensitivity of the planned Square Kilometer Array [Furlanetto and Briggs, 2004], and thus will not be feasible until at least 2020. Statistical observations of the fluctuation power spectrum, on the other hand, should be obtainable with much smaller radio telescope arrays since statistical measurements allow a greater degree of information compression, thus increasing the effective signal to noise ratio in the measurements. Characterizing the power spectrum and its evolution would provide a wealth of information about structure formation and the fundamental astrophysics behind reionization. In large part because of this promise of opening the Dark Ages to scrutiny, as well as because of new enabling technologies, several radio-frequency experiments are underway that hope to detect the redshifted 21 cm background produced by neutral hydrogen above z = 6 and constrain its statistical properties.

Links to Reionization and Redshifted 21 cm Projects and Web Pages

* The text on this page is adapted from the introduction of my Ph. D. thesis (pdf).

Return to homepage