21cm tomography


During the epoch of reionization (EOR), starlight from early galaxies created a patchwork quilt of ionized and neutral zones in the intergalactic hydrogen gas. These neutral hydrogen atoms at high redshift contribute a diffuse background of redshifted 21cm radiation
which encodes information about the physical conditions in the early Universe at z>6 during and before the EOR. Tomography of this 21cm background has emerged as a promising probe of cosmology. I have worked on the forecast of the accuracy with which 21cm tomography can constrain cosmological parameters in Mao et al. (2008), and constrain inflationary parameters in Barger et al. (2009). In principle, 21cm tomography has arguably greater long-term potential than the cosmic microwave background (CMB), because the former is able to map most of our horizon volume while the latter mainly probes a thin shell from z≈1100 (the information garnered about cosmological parameters grows with the volume mapped; see the figure on the right).  Our result shows that a futuristic square-kilometer interferometer array, the Omniscope (Tegmark & Zaldarriaga 2009, 2010), can improve the sensitivity to spatial curvature, neutrino masses, and spectral index running to unprecedentedly high precision (∆Ωk≈0.0002, ∆mν≈0.007eV, ∆α≈0.0002). Much of this research has been done in collaboration with Max Tegmark, Matt McQuinn, Oliver Zahn, Matias Zaldarriaga, Yu Gao, Vernon Barger, and Danny Marfatia.

Since only neutral hydrogen atoms can emit 21cm radiation, the cosmic 21cm background is a direct and sensitive probe of the EOR. I have been involved in numerically simulating the cosmic 21cm signal and its statistics, using large-scale N-body simulations combined with cosmological radiative transfer simulations, in order to investigate whether 21cm tomography can diagnose which types of galaxies are most responsible for cosmic reionization. In Iliev et al. (2011), we find that 21cm power spectra observed by the first generation of 21cm interferometer arrays (e.g. MWA) are able to distinguish the Universe ionized only by HMACHs (high-mass atomic-cooling halos, M ≥ 109 M) alone from that by both HMACHs and LMACHs (low-mass atomic-cooling halos, 108 ≤ M ≤ 109 M) together.


In view of this promise which observations of 21cm power spectra hold for testing and constraining cosmological and astrophysical models to high precision, further progress is required to ensure that predictions are accurate enough to fulfill this promise. This accuracy depends not only on the realistic astrophysical modeling of reionization and neutral hydrogen spin temperature, but also on the accuracy of the methods used to extract the 21cm signal from simulations of the EOR. While my ongoing project is to push forward the progress of the former (see a little detail below), in Mao et al. (2011) I have worked on the latter (i.e. 21cm methodology) issue to take full account of (nonlinear) effects of peculiar velocity on the 21cm signal. We investigate the proper account of the redshift-space distortion for 21cm brightness temperature with or without the optically-thin approximation, to avoid the diverging brightness temperature that appears in the optically-thin approximation. We then set out to build solid and self-consistent computational schemes to predict the fully nonlinear 21cm signal accurately in observer redshift-space, given density, velocity and ionization fraction information in real space. Along the line, we propose two numerical schemes that incorporate peculiar velocity in the optically-thin approximation accurately. One is particle-based (“PPM-RRM”), the other grid-based (“MM-RRM”), and while the former is most accurate, we demonstrate that the latter is computationally more efficient and can be optimized so as to achieve sufficient accuracy.


The figures above give you a flavor of how peculiar velocity changes the three-dimensional 21cm fluctuation power spectra dramatically. The 3-D 21cm power spectrum uncorrected for peculiar velocity (left panel, computed from numerical simulation data) is statistically isotropic. Applying for linear velocity (middle panel) makes it distorted along the line-of-sight (x-axis). Implementing a numerical scheme (the “PPM-RRM” scheme) to take full account of nonlinear velocity (right panel) finds the compressed nature of the large-k modes in power spectrum, too, albeit with some numerical noise.

The movie on the right makes a more quantitative comparison between the angle-averaged 21cm power spectra with and without correction for peculiar velocity, i.e. plots of the ratio of power spectra vs. wavenumber k. As reionization proceeds, the ratio evolves rather nonlinearly, changing both amplitude and shape non-monotonically.


I have been also involved in a project that investigates the light-cone effect on the 21cm power spectra. The light-cone effect results from the evolution of cosmic density and ionization fraction between the front-end and back-end of an observed patch with finite frequency bandwidth. In Datta et al. (2011), we find that the light-cone effect may change the power spectrum up to ∼50% at large scales, while leaving the small-scale powers less affected.

Much of the research above using numerical simulations has been done in collaboration with Paul Shapiro, Ilian Iliev, Garrelt Mellema, Kyungjin Ahn, Ue-Li Pen, Jun Koda, and Kanan Datta.


I am working to develop a new cosmological radiative transfer code that can be used to simulate accurately and efficiently not only the reionization of hydrogen and helium atoms, but also the thermal condition and neutral hydrogen spin temperature of the intergalactic medium (IGM). This code will include and trace all spectra of ionizing photons, i.e. not only those ultraviolet photons that are most responsible for hydrogen reionization, but also those hard photons that heat the IGM gas immediately after first galaxies form and later ionize the helium atoms. The latter was often neglected in current reionization simulations, partly because of the difficulty associated with the very long mean-free-path of hard photons. More details are to be reported. Stay tuned. 


Alternative theories of gravity

The reason that Einstein's theory of General Relativity (GR) is now taken so seriously is that it has passed so many experimental tests. Specifically, it has been embedded in more general classes of theories, the so-called Parametrized Post-Newtonian (PPN) formalism, with various constants that parametrize departures from GR, and these PPN parameters have been observationally constrained to be extremely close to the values predicted by GR. I have further generalized such tests to allow testing assumptions that are normally not questioned, for example whether a type of space-time distortion known as torsion exists, and whether the Einstein field equation contains extra terms that are general functions of Ricci scalar (i.e. f(R) gravity) and could affect cosmic expansion and structure formation. Specifically, torsion is defined as the antisymmetric part of affine connection in a space-time manifold, so in principle it could affect precession of a gyroscope in Earth orbit.  I showed that the satellite experiment Gravity Probe B can measure the values of torsion parameters to the unprecedented accuracy of one part in ten thousand. Much of this research has been done in collaboration with Max Tegmark, Alan Guth, Emory Bunn, Tom Faulkner, and Serkan Cabi.