The BINGO (Baryon Acoustics Oscillations from Integrated Neutral Gas Observations) project is a double dish radio telescope aiming at observation of the 21cm line corresponding to hyperfine interaction of the atomic Hydrogen (parallel versus antiparallel electron and proton spins). It will survey a sky area of ∼ 5000 square degrees in a redshift range from 0.127 to 0.449 (corresponding to a frequency span of 280MHz) and angular resolution about 40 arcmin. The instrument, in its Phase 1, will operate with 28 feed horns and receivers with dual polarization and is designed to have a system temperature of 70K. Moreover, in view of the telescope characteristics, it is able to describe phenomena at very short time scales, being thus a fundamental device to study pulsars and Fast Radio Bursts (FRB).

Our group has mainly focused on theoretical cosmology side of the BINGO project. Together with our collaborators, we have experience in constructing and studying theoretical models, analysing linear and non-linear perturbations, calculating 21-cm power spectra, making maps, performing foreground removal and carrying out cosmological constraints. We are currently participating in several topics of the BINGO pipeline.

Our goal is to use the 21-cm signal and FRBs which will be detected by BINGO to improve the current understanding of our Universe, including its components and history. We are specially interested in understanding the nature of dark energy and whether it has any relation to another mysterious component, dark matter. In this aspect, wee have proposed and analyzed some new dark energy models. We are also implementing numerical codes to calculate their 21-cm angular power spectra, make forecasts and analyse the real data when available. Moreover, our group is also working on foreground removal techniques, which is the main contaminant of 21-cm cosmology.

The BINGO project is led by the University of Sao Paulo in Brazil, with the participation of teams from the University of Manchester, University College London, the Swiss Federal Institute of Technology and China. The Chinese team is led by co-chief scientist Professor Bin Wang from the Center for Gravitation and Cosmology (CGC) at Yangzhou University, consisting of Shanghai Jiao Tong University, Yangzhou University and CETC as the core participating units. The project was inaugurated on July 6th, 2021 in Shanghai:

In the following we introduce some specific aspects of our research directions in more details:

A) Cosmological models

The ΛCDM model encounters some theoretical and observational difficulties which have led scientists to look for more sophisticated ideas to explain the current accelerated expansion. Simple generalizations consist of deviations to the cosmological constant equation of state, which can be constant or evolving in time. Another interesting possibility is that dark energy and dark matter do not evolve independently, but rather interact with each other (this could explain, e.g. why their energy density is about equal today -- the so-called "coincidence problem"). Our group intends to analyse several different interacting models and examine their signatures in 21cm spectrum. Another way to analyze the interaction is to reconstruct the interacting function directly from the observational data. Furthermore, we are also interested in detecting any deviation from Einstein's general theory of relativity.

B) Foregrounds and Component Separation 

The real map suffers from several contaminants that obscures the 21-cm signal. The most serious one is the foreground which is several orders of magnitude higher than the desired signal. Because of this, we need to develop techniques of foreground removal. In fact, an experiment like BINGO should not rely on only one technique, instead we must compare the results from different methods to check their robustness. With recent trends and heavy developments, machine-learning-based techniques are viable options for BINGO science.

Besides foregrounds, it is very important to characterize the map noise, which consists of gain fluctuations, detector 1/f noise, atmospheric contamination, and radio frequency interference from the ground. Numerical simulations incorporating general antenna design and configuration will be established to understand these noise components and provide guidance for noise bias reduction. Furthermore, mock noise realizations will be generated from these numerical simulations for power spectrum error estimation, detectability of the BAO signature, and pipeline validation.

As an indispensable step of the power spectrum analysis, the beam profile will be required to deconvolve the measured 21-cm maps. We will employ state-of-the-art electromagnetic simulations to precisely characterize the beam profile, comparing the simulated beam pattern with lunar-transit observations and near-field transmitter measurements and fine-tuning the beam models. Moreover, we will estimate the non-Gaussianity of the beam and investigate its systematic effects and possible biases for power spectrum analysis.

C) HI Signal and Cosmological Constraints

In order to test our cosmological models and constrain their parameters, we need precise predictions of what to expect from 21-cm measurements. Therefore, we have studied the 21-cm power spectra considering linear and non-linear effects.

The purpose of preparing BINGO pipeline is to include all the above mentioned effects and be capable to extract the 21-cm signal and put constraints on our theoretical models. Before the real data is available, we are working on forecasting the expected results and also preparing the tools. At the moment we have a Fisher matrix code, which allows us to make an optimal forecasts for a HI experiment. The next step is to develop the codes to calculate the likelihood and Markov Chain Monte Carlo parameter estimator, which allow a more realistic exploration of the parameter space.

D) Primordial Power Spectrum

HI intensity mapping surveys could also be used in combination with CMB experiments to constrain the primordial scalar spectral index (ns) and its runnings, and test predictions of popular single-field slow-roll inflation models.

Moreover, we can study non-Gaussianities using either a scale-dependent bias of the power spectrum or the bispectrum. Previous forecasts show that even in the presence of foreground contamination, upcoming HI intensity mapping observations of the large-scale structure could put tight constraints on featured models, potentially improving over the two-dimensional CMB measurements.

E) Cosmological constraints from FRB

FRB is a transient radio pulse, they are brief and bright. The astrophysical origin of FRB remains inconclusive. In general, the radio astronomy data gives the time series of different frequency channels and finally gives an image of frequency and time. The FRB in the high frequency channel arrives earlier than the low frequency. FRB has a large dispersion of frequency. That means the arrival time of different frequency channels are different. The amount of dispersion is quantified by the time delay of the pulse. The dispersion measure (DM) is defined as the integral of the electron number density. By measuring the time delay of two frequency channels, we can directly get DM.

Because of the large dispersion measure, FRBs are thought to be extragalactic. In general, the total dispersion measure is from three contributions, including Milky Way (MW), intergalactic medium (IGM) and host galaxy (HG). Usually, DMMW are well measured and DMHG is poorly known. We define DME to represent the contributions out of Milky Way. The largest contribution of dispersion measure is from the intergalactic medium. The average of DMIGM is related to the cosmological models. FRB measurements do not introduce additional parameters. If we know the redshift of FRB, dispersion measure can be used to constrain cosmological parameters. Up to now, 129 FRBs have been reported by 9 Telescopes. However, there are redshift measurements for only a few of them.

In order to forecast the constraints on cosmological parameters using the FRB’s dispersion measure detected by BINGO, we are building a mock catalog of observed FRBs. First, we generate the intrinsic FRB population. After some selection effects (for example, the luminosity, pulse width, beam response, etc.) from BINGO, we get the simulated population. Then we need to compare the simulated population with the real FRBs. Now we have simulated the FRBs observed by BINGO for 1000 days, including 862 mock data. We have obtained the dispersion measure from Milky Way, intergalactic medium and host galaxy, as well as the redshifts, location and some other parameters related to telescope of each FRB. From these mock data, we can analyze the distribution, mean value, error and so on. Then, we find the distribution of DM and DMMW are consistent with the real data from FRB catalogue. Further, we will test the ability of FRB to constrain cosmological parameters, and then combine FRB measurement with the latest result from Planck to see how it can improve cosmological parameters’ constraints.