Research

Cosmology is the main area of my research. In particular, I want to understand various unknown facts during the onset of the first sources of light (often termed as ‘cosmic dawn’) and the epoch of reionization (EoR), when the neutral hydrogen (HI ) in the intergalactic medium (IGM) was ionized in the Universe. Recently made progress in detecting hundreds of galaxies and dozens of quasars above redshift 6, has revolutionized our understanding of the early sources in the Universe. Still, we are unable to answer questions like: when did the first sources form in the universe?, what are their properties?, how did IGM evolve during the EoR etc.

Observations of the redshifted HI 21-cm signal from the IGM are a powerful probe of this epoch and can answer many such questions. Thus a huge effort is in place to measure the redshifted 21-cm signal from the cosmic dawn and EoR by low-frequency radio telescopes like LOFAR, MWA, PAPER, GMRT, SKA etc. Also, many theoretical works using analytical, numerical and semi-numerical methods are going on. These works will be helpful in understanding the observational results and designing new observational strategies.

My scientific contributions towards this have been developing different tools for modelling and studying various aspects of the HI signal from the EoR. A brief description of my research is given below.

Modelling & studying different aspects of the HI signal from the EoR

I developed a semi-numerical code GRIZZLY (arXiv:1406.4157) which generates 3D maps of the HI signal from the EoR. It uses a list of dark matter halos, density, and velocity fields from an N-body simulation and a one-dimensional radiative transfer scheme to model the H I signal. In addition to the ionization fluctuations, it also accounts for the spin temperature fluctuations arising from the inhomogeneous X-ray heating and Lyα coupling during the EoR (see Fig. 1). We studied the impact of different types of sources such as galaxies, high-mass X-ray binaries, and mini-quasars on the H I signal (arXiv:1607.02779, arXiv:1905.10386, arXiv:2012.11616) using GRIZZLY . We also studied the impact of the line-of-sight effects such as redshift-space distortion (RSD) and the light-cone (LC) effects on the HI signal (arXiv:1504.05601) using GRIZZLY . We used the same code to interpret different radio observations as well as develop the EoR radio observation strategies (arXiv:1909.12317).

Alternative to GRIZZLY , I also developed an analytical code to model the EoR HI signal in terms of the statistical quantities such as the volume-averaged signal, power spectrum (PS), the size distribution of the ionized regions, etc. Using the same analytical code, we studied the impact of Fuzzy dark matter (FDM) particles on the EoR HI signal. We obtained strong constraints on the light scalar FDM particle mass using EDGES HI signal measurements (arXiv:1812.09760). We have also used the same code and the EDGES observations to constrain a radio emission model from the decay of unstable particles.

Figure 1: A light cone showing the redshift (z) evolution of the brightness temperature of the HI signal generated using GRIZZLY for a set of simulation parameters.

Movie 1: A movie showing the evolution of the brightness temperature of HI 21-cm signal from the Cosmic Dawn and Epoch of Reionization. The movie is generated using simulation maps from GRIZZLY code.

Movie 1 shows the evolution of the 21-cm signal through redshift or time. With the formation of the first stars, the cosmic Dawn began (around redshift 20, in this case). The Lyα, X-ray and radio background determine the strength of the signal during this time as the gas in the IGM remained mostly neutral. X-rays penetrate deep into the IGM and heat the gas. This phase is followed by an era where ionization of the gas in the IGM by the ultraviolet photons from the first galaxies became important. Individual ionized regions around the sources started to overlap and complete the reionization process around redshift 6.

Interpretation of the EoR H I 21-cm observations

Extracting astrophysical and cosmological information from the HI 21-cm observations is not straightforward as, in addition to the cosmological dependence, the characteristics of the expected signal depend crucially on specific properties of the early sources and their redshift evolution. Because of this complexity, an exploration of many theoretical models of the expected 21-cm signal is necessary to interpret the results from radio observations. Recently, we have developed a Bayesian inference framework that uses the results from GRIZZLY to constrain the physical states of the IGM besides the properties of the sources formed during the EoR. This framework constrains the thermal and the ionized states of the IGM during the EoR in terms of quantities such as average ionization fraction, the average volume fraction of ‘heated regions’, i.e., regions with temperature larger than the CMB temperature and size distribution of the heated regions. Previously, we have applied this framework to interpret the EoR observations from the interferometers such as LOFAR, and MWA (arXiv:2002.07195, 2103.07483).

Figure shows upper limits on the spherically averaged power spectrum of the 21-cm signal from redshift 9.1. This limit is obtained using 10 nights of observation with the LOFAR interferometer (Mertens et al 2020).

Figure shows constraints on different quantities that characterize the IGM at redshift 9.1. The limits are obtained using GRIZZLY simulations (Ghara et al 2020)

Developing strategies for detecting the EoR HI signal

One of the main challenges of the EoR HI experiments is to remove the contribution of the galactic and extra-galactic foregrounds which are stronger than the expected HI signal by several orders of magnitude. Their smooth frequency dependence should allow separation from the fluctuating HI signal, allowing them to be either subtracted, suppressed, or avoided. Moreover, a long observation time is necessary to reduce the instrumental noise. Thus, it is necessary to design optimum observation strategies for detecting the HI signal from the EoR.

I have studied different strategies to detect the signal using mock images (arXiv:1607.02779, 1801.06550), and visibilities (arXiv:1511.07448) in realistic observation setups. Recently, I also developed a Bayesian inference framework (arXiv:1909.12317) which uses matched filters to enhance the detectability of the signal. It also efficiently finds the position, size, and amplitude of the strongest HI signal in the field of view.

The figure on the left shows a simulated 21-cm map that contains a bright quasar. The ionized region around the quasar is close to a spherical shape. The use of a spherical filter in the matched-filtering technique will increase the detectability of such a large ionized region.

Page updated

Google Sites

Report abuse