Signals from the Universe’s First Stars Detectable with New Radio Telescope Strategy

Researchers investigating the earliest epochs of the universe focus on the 21cm signal emitted by neutral hydrogen as a primary means of probing the Dark Ages. Shintaro Yoshiura from Mizusawa VLBI Observatory, Fumiya Okamatsu from Nihon University, and Tomo Takahashi from Saga University, et al. present a detailed assessment of how readily this faint signal can be detected and used to distinguish between different cosmological models. Their work demonstrates that wide-band observations spanning 1-50MHz, coupled with sufficiently low error levels, can reveal evidence of a 21cm signal, excluding models with particularly smooth spectral characteristics. Crucially, the study highlights the importance of observations at frequencies below 15MHz to mitigate foreground contamination and confirms that even limited frequency resolution can capture the intrinsic shape of the 21cm signal, offering a pathway to unlock the secrets of the cosmos’ infancy.

Detecting cosmological signals within the 21cm epoch of reionisation using Bayesian evidence requires careful foreground mitigation and noise modelling

Scientists are now leveraging observations of the 21cm signal from neutral hydrogen to probe the very first moments of the Universe, known as the Dark Ages. This faint signal, detected at radio frequencies below 50MHz, offers a unique window into cosmology, as theoretical models predict a distinct spectral shape for it.
Recent work assesses the feasibility of detecting this signal and differentiating between various cosmological models, considering realistic foreground interference and potential observing strategies. Through a Bayesian evidence-based comparison, researchers demonstrate that wide-band observations spanning 1, 50MHz can identify non-zero 21cm signals from most models examined, with the exception of one exhibiting a smooth spectrum peaking at lower frequencies.

Crucially, observations below 15MHz are identified as essential for mitigating the effects of foreground contamination. Even with measurements taken at 5MHz intervals across the 1, 50MHz range, the 21cm signal remains detectable provided error levels are sufficiently low. This finding suggests that the intrinsic spectral shape of the signal can be accurately captured even with a limited number of frequency channels, representing a significant step forward in observational cosmology.
The study highlights the potential to unlock insights into the early Universe without being hampered by limitations in data resolution. This research employed a Bayesian evidence approach to evaluate the detectability and model-selection capabilities of 21cm observations. Several cosmological models were tested against physically motivated foregrounds, optimistic error levels, and various observing strategies.

The analysis revealed that wide-band observations are capable of identifying non-zero 21cm signals, with the exception of a specific model characterized by a low-frequency peak. The importance of low-frequency observations, below 15MHz, was confirmed, as they are critical for minimizing interference from foreground emissions.

Furthermore, the study demonstrates that even with relatively sparse frequency sampling, 5MHz intervals, the 21cm signal can be identified if instrumental errors are minimized. This suggests that capturing the essential spectral information does not necessarily require extremely high-resolution data, simplifying the demands on future instrumentation. These findings pave the way for upcoming projects, including moon-orbiting and lunar-based missions, designed to detect and characterize the 21cm signal from the Dark Ages and unlock fundamental knowledge about the early Universe.

Bayesian model selection using PolyChord and 21cm signal simulations from modified RECFAST code allows for robust cosmological parameter estimation

A Bayesian evidence-based comparison forms the core of this research, assessing the detectability of the 21cm signal from the Dark Ages and differentiating between various cosmological models. The study utilizes PolyChord to perform nested sampling, a technique employed to evaluate the Bayesian evidence, denoted as Zi, for each model.

This allows for the calculation of the Bayes factor, ∆ln Zi,j, quantifying the relative likelihood of different models based on observational data. Values of ∆ln Zi,j less than 1 indicate negligible evidence, while values between 1 and 3, 3 and 5, and greater than 5 represent positive, strong, and very strong evidence, respectively.

Eight distinct cosmological models for the 21cm signal were employed, alongside a null hypothesis model with no 21cm signal, all calculated using modifications to the RECFAST code. These models, ΛCDM, DMBw, DMBs, EDE, ERB, LDMD, PMFw, and PMFs, were designed to explore variations in the 21cm signal’s spectral shape, with peak amplitudes and frequencies summarized in Table I.

Cosmological parameters were fixed at Ωbh2 = 0.02237, Ωmh2 = 0.14237, Yp = 0.2436 and H0 = 67.36km s−1 Mpc−1 to focus on the intrinsic 21cm spectral shape. The research investigated observing strategies covering the 1-50MHz frequency range, with measurements taken at 5MHz intervals, to determine the minimum error levels required for signal identification.

The 21cm signal itself is defined as proportional to 1 −Tγ/TS, where Tγ represents the radio background temperature and TS is the spin temperature. Radio background temperature was calculated as TR = Tcmb + TER (ν/ν78)−2.6, with TER set to 0.03 K for the ERB model and TR equal to Tcmb for all other models. This methodology enabled the study to demonstrate that wide-band observations, particularly below 15MHz, are crucial for avoiding degeneracy with foreground signals and accurately capturing the 21cm spectral shape even with a limited number of frequency channels.

Cosmological model discrimination via 21cm signal detection between 1 and 50MHz is exceptionally challenging

Observations covering frequencies between 1 and 50MHz can identify evidence of non-zero 21cm signals from several cosmological models, excluding only the model featuring a smooth spectrum peaking at lower frequencies. Wide-band observations are crucial for probing the Dark Ages and utilizing the 21cm signal from hydrogen as a cosmological tool, as the standard cosmological model predicts a well-defined spectral shape at these frequencies.

Specifically, data collected below 15MHz are essential to mitigate degeneracies with foreground emissions during analysis. Even with measurements taken at 5MHz intervals across the 1-50MHz range, the 21cm signal remains identifiable provided error levels are sufficiently low. This demonstrates that the intrinsic spectral shape of the 21cm signal can be accurately captured without requiring an excessive number of frequency channels.

Eight distinct 21cm signal models were employed in this study, alongside a no-signal model, to assess detectability and model-selection capabilities. The ΛCDM model, serving as a baseline, exhibits an absorption trough centered at 16MHz with an amplitude of -40.1mK. The DMBw model demonstrates a stronger absorption signal, reaching -62.1mK at 17MHz, while the DMBs model shows an even more pronounced absorption at -160.6mK, peaking at 14MHz.

The EDE model presents an absorption trough of -57.9mK at 14MHz, and the LDMD model exhibits a weaker absorption signal with a peak of -29.8mK at 16MHz. Conversely, the PMFw model shows a shallower absorption trough of -21.3mK at 17MHz, and the PMFs model uniquely displays an emission line, peaking at 25.7mK at 15MHz. Table I summarizes these spectral features, detailing the peak amplitude and frequency for each model, excluding the ERB model which does not exhibit an absorption trough.

Detecting cosmological signatures within the 21cm signal using low-frequency radio observations is a challenging but promising avenue for understanding the early universe

Observations of the 21cm signal from neutral hydrogen offer a unique means of probing the cosmological Dark Ages. This work assesses the feasibility of detecting this signal and differentiating between various cosmological models using physically realistic foreground estimations and anticipated observational constraints.

Analyses demonstrate that wide-band observations spanning frequencies from 1 to 50MHz can reliably identify non-zero 21cm signals from the models examined, with the exception of those exhibiting smooth spectra peaking at lower frequencies. Crucially, observations at frequencies below 15MHz are vital to mitigate the effects of foreground contamination and avoid model degeneracy.

Even with relatively limited frequency resolution, measurements taken at 5MHz intervals, a discernible 21cm signal can be detected provided the observational errors are sufficiently small. This suggests that the intrinsic spectral shape of the 21cm signal can be accurately determined without requiring an excessively fine frequency sampling.

Different cosmological models produce distinct 21cm signal characteristics, with some exhibiting strong absorption and others displaying emission features. The authors acknowledge limitations stemming from computational expense, preventing a full exploration of parameter spaces for all models through nested sampling techniques.

Furthermore, accurately modelling instrumental systematic errors, such as calibration inaccuracies and ionospheric effects, remains a significant challenge. Future research should focus on refining foreground modelling and developing strategies to mitigate instrumental systematic errors to fully realise the potential of 21cm cosmology. These findings establish a clear path toward utilising 21cm observations to constrain cosmological parameters and deepen our understanding of the early universe.