Search

Neural network (NN) emulators of the global 21 cm signal need emulation error much less than the observational noise in order to be used to perform unbiased Bayesian parameter inference. To this end, we introduce $\texttt{21cmLSTM}$ -- a long short-term memory (LSTM) NN emulator of the global 21 cm signal that leverages the intrinsic correlation between frequency channels to achieve exceptional accuracy compared to previous emulators, which are all feedforward, fully connected NNs. LSTM NNs are a type of recurrent NN designed to capture long-term dependencies in sequential data. When trained and tested on the same simulated set of global 21 cm signals as the best previous emulators, $\texttt{21cmLSTM}$ has average relative rms error of 0.22% -- equivalently 0.39 mK -- and comparably fast evaluation time. We perform seven-dimensional Bayesian parameter estimation analyses using $\texttt{21cmLSTM}$ to fit global 21 cm signal mock data with different adopted observational noise levels, $\sigma_{21}$. The posterior $1\sigma$ rms error is $\approx3\times$ less than $\sigma_{21}$ for each fit and consistently decreases for tighter noise levels, showing that $\texttt{21cmLSTM}$ can sufficiently exploit even very optimistic measurements of the global 21 cm signal. We made the emulator, code, and data sets publicly available so that $\texttt{21cmLSTM}$ can be independently tested and used to retrain and constrain other 21 cm models., Comment: 15 pages, 8 figures, 3 tables. Accepted by ApJ

Although neural-network-based emulators enable efficient parameter estimation in 21-cm cosmology, the accuracy of such constraints is poorly understood. We employ nested sampling to fit mock data of the global 21-cm signal and high-$z$ galaxy ultraviolet luminosity function (UVLF) and compare for the first time the emulated posteriors obtained using the global signal emulator ${\tt globalemu}$ to the `true' posteriors obtained using the full model on which the emulator is trained using ${\tt ARES}$. Of the eight model parameters we employ, four control the star formation efficiency (SFE), and thus can be constrained by UVLF data, while the remaining four control UV and X-ray photon production, and the minimum virial temperature of star-forming halos ($T_{\rm min}$), and thus are uniquely probed by reionization and 21-cm measurements. For noise levels of 50 and 250 mK in the 21-cm data being jointly-fit, the emulated and `true' posteriors are consistent to within $1\sigma$. However, at lower noise levels of 10 and 25 mK, ${\tt globalemu}$ overpredicts $T_{\rm min}$ and underpredicts $\gamma_{\rm lo}$, an SFE parameter, by $\approx3-4\sigma$, while the `true' ${\tt ARES}$ posteriors capture their fiducial values within $1\sigma$. We find that jointly-fitting the mock UVLF and 21-cm data significantly improves constraints on the SFE parameters by breaking degeneracies in the ${\tt ARES}$ parameter space. Our results demonstrate the astrophysical constraints that can be expected for global 21-cm experiments for a range of noise levels from pessimistic to optimistic, and also the potential for probing redshift evolution of SFE parameters by including UVLF data., Comment: 19 pages, 8 figures, 2 tables. Accepted by ApJ

One of the last unexplored windows to the cosmos, the Dark Ages and Cosmic Dawn, can be opened using a simple low frequency radio telescope from the stable, quiet lunar farside to measure the Global 21-cm spectrum. This frontier remains an enormous gap in our knowledge of the Universe. Standard models of physics and cosmology are untested during this critical epoch. The messenger of information about this period is the 1420 MHz (21-cm) radiation from the hyperfine transition of neutral hydrogen, Doppler-shifted to low radio astronomy frequencies by the expansion of the Universe. The Global 21-cm spectrum uniquely probes the cosmological model during the Dark Ages plus the evolving astrophysics during Cosmic Dawn, yielding constraints on the first stars, on accreting black holes, and on exotic physics such as dark matter-baryon interactions. A single low frequency radio telescope can measure the Global spectrum between ~10-110 MHz because of the ubiquity of neutral hydrogen. Precise characterizations of the telescope and its surroundings are required to detect this weak, isotropic emission of hydrogen amidst the bright foreground Galactic radiation. We describe how two antennas will permit observations over the full frequency band: a pair of orthogonal wire antennas and a 0.3-m$^3$ patch antenna. A four-channel correlation spectropolarimeter forms the core of the detector electronics. Technology challenges include advanced calibration techniques to disentangle covariances between a bright foreground and a weak 21-cm signal, using techniques similar to those for the CMB, thermal management for temperature swings of >250C, and efficient power to allow operations through a two-week lunar night. This simple telescope sets the stage for a lunar farside interferometric array to measure the Dark Ages power spectrum.

The Dark Ages, probed by the redshifted 21-cm signal, is the ideal epoch for a new rigorous test of the standard LCDM cosmological model. Divergences from that model would indicate new physics, such as dark matter decay (heating) or baryonic cooling beyond that expected from adiabatic expansion of the Universe. In the early Universe, most of the baryonic matter was in the form of neutral hydrogen (HI), detectable via its ground state's spin-flip transition. A measurement of the redshifted 21-cm spectrum maps the history of the HI gas through the Dark Ages and Cosmic Dawn and up to the Epoch of Reionization (EoR). The Experiment to Detect the Global EoR Signature (EDGES) recently reported an absorption trough at 78 MHz (redshift z of 17), similar in frequency to expectations for Cosmic Dawn, but about 3 times deeper than was thought possible from standard cosmology and adiabatic cooling of HI. Interactions between baryons and slightly-charged dark matter particles with electron-like mass provide a potential explanation of this difference but other cooling mechanisms are also being investigated to explain these results. The Cosmic Dawn trough is affected by cosmology and the complex astrophysical history of the first luminous objects. Another trough is expected during the Dark Ages, prior to the formation of the first stars and thus determined entirely by cosmological phenomena (including dark matter). Observations on or in orbit above the Moon's farside can investigate this pristine epoch (15-40 MHz; z=100-35), which is inaccessible from Earth. A single cross-dipole antenna or compact array can measure the amplitude of the 21-cm spectrum to the level required to distinguish (at >5$\sigma$}) the standard cosmological model from that of additional cooling derived from current EDGES results. This observation constitutes a powerful, clean probe of exotic physics in the Dark Ages., Comment: 7 pages, 6 figures, Science whitepaper submitted to the Astro2020 Decadal Survey