Numerous studies have established the theoretical potential of the 21cm bispectrum to boost our understanding of the Epoch of Reionization (EoR), and therefore early generation of stars and galaxies. In this paper we look at the impact of foregrounds and instrumental effects on the 21cm bispectrum and our ability to measure it. Unlike the power spectrum for which (in the absence of instrumental effects) there is a window clear of smooth-spectrum foregrounds in which the 21cm power spectrum may be detectable, there is no such "EoR window" for the bispectrum. Instead, on smaller scales EoR structure modulates that of the foregrounds (FG) to alter the EoR+FG bispectrum from that of the foregrounds in a complex manner. On larger scales the EoR structures are completely swamped by those of the foregrounds, and the EoR+FG bispectrum is entirely dominated by that of the foregrounds. It is therefore unlikely that the bispectrum will be useful in cases where we are observationally restricted to using foreground avoidance techniques. We also find that there is potential for instrumental effects to seriously corrupt the bispectrum, possibly even rendering the bispectrum useless for parameter estimation. On larger scales, foreground removal using GMCA is found to recover the EoR bispectrum to a reasonable level of accuracy (even better than the power spectrum for certain configurations). Further studies are necessary to understand the error and/or bias associated with foreground removal before the 21cm bispectrum can be practically applied in analysis of future data.

We compute the bispectra of the 21cm signal during the Epoch of Reionization for three different reionization scenarios that are based on a dark matter N-body simulation combined with a self-consistent, semi-numerical model of galaxy evolution and reionization. Our reionization scenarios differ in their trends of ionizing escape fractions ($f_\mathrm{esc}$) with the underlying galaxy properties and cover the physically plausible range, i.e. $f_\mathrm{esc}$ effectively decreasing, being constant, or increasing with halo mass. We find the 21cm bispectrum to be sensitive to the resulting ionization topologies that significantly differ in their size distribution of ionized and neutral regions throughout reionization. From squeezed to stretched triangles, the 21cm bispectra features a change of sign from negative to positive values, with ionized and neutral regions representing below-average and above-average concentrations contributing negatively and positively, respectively. The position of the change of sign provides a tracer of the size distribution of the ionized and neutral regions, and allows us to identify three major regimes that the 21cm bispectrum undergoes during reionization. In particular the regime during the early stages of reionization, where the 21cm bispectrum tracks the peak of the size distribution of the ionized regions, provides exciting prospects for pinning down reionization with the forthcoming Square Kilometre Array.

We apply two methods to estimate the 21~cm bispectrum from data taken within the Epoch of Reionisation (EoR) project of the Murchison Widefield Array (MWA). Using data acquired with the Phase II compact array allows a direct bispectrum estimate to be undertaken on the multiple redundantly-spaced triangles of antenna tiles, as well as an estimate based on data gridded to the $uv$-plane. The direct and gridded bispectrum estimators are applied to 21 hours of high-band (167--197~MHz; $z$=6.2--7.5) data from the 2016 and 2017 observing seasons. Analytic predictions for the bispectrum bias and variance for point source foregrounds are derived. We compare the output of these approaches, the foreground contribution to the signal, and future prospects for measuring the bispectra with redundant and non-redundant arrays. We find that some triangle configurations yield bispectrum estimates that are consistent with the expected noise level after 10 hours, while equilateral configurations are strongly foreground-dominated. Careful choice of triangle configurations may be made to reduce foreground bias that hinders power spectrum estimators, and the 21~cm bispectrum may be accessible in less time than the 21~cm power spectrum for some wave modes, with detections in hundreds of hours.

We present analysis of the normalised 21-cm bispectrum from fully-numerical simulations of intergalactic-medium heating by stellar sources and high-mass X-ray binaries (HMXB) during the cosmic dawn. Lyman-$\alpha$ coupling is assumed to be saturated, we therefore probe the nature of non-Gaussianities produced by X-ray heating processes. We find the evolution of the normalised bispectrum to be very different from that of the power spectrum. It exhibits a turnover whose peak moves from large to small scales with decreasing redshift, and corresponds to the typical separation of emission regions. This characteristic scale reduces as more and more regions move into emission with time. Ultimately, small-scale fluctuations within heated regions come to dominate the normalised bispectrum, which at the end of the simulation is almost entirely driven by fluctuations in the density field. To establish how generic the qualitative evolution of the normalised bispectrum we see in the stellar + HMXB simulation is, we examine several other simulations - two fully-numerical simulations that include QSO sources, and two with contrasting source properties produced with the semi-numerical simulation 21cmFAST. We find the qualitative evolution of the normalised bispectrum during X-ray heating to be generic, unless the sources of X-rays are, as with QSOs, less numerous and so exhibit more distinct isolated heated profiles. Assuming mitigation of foreground and instrumental effects are ultimately effective, we find that we should be sensitive to the normalised bispectrum during the epoch of heating, so long as the spin temperature has not saturated by $z \approx 19$.

Upcoming experiments such as the SKA will provide huge quantities of data. Fast modelling of the high-redshift 21cm signal will be crucial for efficiently comparing these data sets with theory. The most detailed theoretical predictions currently come from numerical simulations and from faster but less accurate semi-numerical simulations. Recently, machine learning techniques have been proposed to emulate the behaviour of these semi-numerical simulations with drastically reduced time and computing cost. We compare the viability of five such machine learning techniques for emulating the 21cm power spectrum of the publicly-available code SimFast21. Our best emulator is a multilayer perceptron with three hidden layers, reproducing SimFast21 power spectra $10^8$ times faster than the simulation with 4% mean squared error averaged across all redshifts and input parameters. The other techniques (interpolation, Gaussian processes regression, and support vector machine) have slower prediction times and worse prediction accuracy than the multilayer perceptron. All our emulators can make predictions at any redshift and scale, which gives more flexible predictions but results in significantly worse prediction accuracy at lower redshifts. We then present a proof-of-concept technique for mapping between two different simulations, exploiting our best emulator's fast prediction speed. We demonstrate this technique to find a mapping between SimFast21 and another publicly-available code 21cmFAST. We observe a noticeable offset between the simulations for some regions of the input space. Such techniques could potentially be used as a bridge between fast semi-numerical simulations and accurate numerical radiative transfer simulations.

The epoch of reionization (EoR) 21-cm signal is expected to be highly non-Gaussian in nature and this non-Gaussianity is also expected to evolve with the progressing state of reionization. Therefore the signal will be correlated between different Fourier modes ($k$). The power spectrum will not be able capture this correlation in the signal. We use a higher-order estimator -- the bispectrum -- to quantify this evolving non-Gaussianity. We study the bispectrum using an ensemble of simulated 21-cm signal and with a large variety of $k$ triangles. We observe two competing sources driving the non-Gaussianity in the signal: fluctuations in the neutral fraction ($x_{\rm HI}$) field and fluctuations in the matter density field. We find that the non-Gaussian contribution from these two sources vary, depending on the stage of reionization and on which $k$ modes are being studied. We show that the sign of the bispectrum works as a unique marker to identify which among these two components is driving the non-Gaussianity. We propose that the sign change in the bispectrum, when plotted as a function of triangle configuration $\cos{\theta}$ and at a certain stage of the EoR can be used as a confirmative test for the detection of the 21-cm signal. We also propose a new consolidated way to visualize the signal evolution (with evolving $\overline{x}_{\rm HI}$ or redshift), through the trajectories of the signal in a power spectrum and equilateral bispectrum i.e. $P(k)-B(k, k, k)$ space.

In this paper we establish the accuracy and robustness of a fast estimator for the bispectrum - the "FFT bispectrum estimator". The implementation of the estimator presented here offers speed and simplicity benefits over a direct sampling approach. We also generalise the derivation so it may be easily be applied to any order polyspectra, such as the trispectrum, with the cost of only a handful of FFTs. All lower order statistics can also be calculated simultaneously for little extra cost. To test the estimator we make use of a non-linear density field, and for a more strongly non-Gaussian test case we use a toy-model of reionization in which ionized bubbles at a given redshift are all of equal size and are randomly distributed. Our tests find that the FFT estimator remains accurate over a wide range of k, and so should be extremely useful for analysis of 21-cm observations. The speed of the FFT bispectrum estimator makes it suitable for sampling applications, such as Bayesian inference. The algorithm we describe should prove valuable in the analysis of simulations and observations, and whilst we apply it within the field of cosmology, this estimator is useful in any field that deals with non-Gaussian data.

This paper considers the impact of Lyman-alpha coupling and X-ray heating on the 21-cm brightness-temperature one-point statistics (as predicted by semi-numerical simulations). The X-ray production efficiency is varied over four orders of magnitude and the hardness of the X-ray spectrum is varied from that predicted for high-mass X-ray binaries, to the softer spectrum expected from the hot inter-stellar medium. We find peaks in the redshift evolution of both the variance and skewness associated with the efficiency of X-ray production. The amplitude of the variance is also sensitive to the hardness of the X-ray SED. We find that the relative timing of the coupling and heating phases can be inferred from the redshift extent of a plateau that connects a peak in the variance's evolution associated with Lyman-alpha coupling to the heating peak. Importantly, we find that late X-ray heating would seriously hamper our ability to constrain reionization with the variance. Late X-ray heating also qualitatively alters the evolution of the skewness, providing a clean way to constrain such models. If foregrounds can be removed, we find that LOFAR, MWA and PAPER could constrain reionization and late X-ray heating models with the variance. We find that HERA and SKA (phase 1) will be able to constrain both reionization and heating by measuring the variance using foreground-avoidance techniques. If foregrounds can be removed they will also be able to constrain the nature of Lyman-alpha coupling.

We investigate the impact of sinks of ionizing radiation on the reionization-era 21-cm signal, focusing on 1-point statistics. We consider sinks in both the intergalactic medium and inside galaxies. At a fixed filling factor of HII regions, sinks will have two main effects on the 21-cm morphology: (i) as inhomogeneous absorbers of ionizing photons they result in smaller and more widespread cosmic HII patches; and (ii) as reservoirs of neutral gas they contribute a non-zero 21-cm signal in otherwise ionized regions. Both effects damp the contrast between neutral and ionized patches during reionization, making detection of the epoch of reionization with 21-cm interferometry more challenging. Here we systematically investigate these effects using the latest semi-numerical simulations. We find that sinks dramatically suppress the peak in the redshift evolution of the variance, corresponding to the midpoint of reionization. As previously predicted, skewness changes sign at midpoint, but the fluctuations in the residual HI suppress a late-time rise. Furthermore, large levels of residual HI dramatically alter the evolution of the variance, skewness and power spectrum from that seen at lower levels. In general, the evolution of the large-scale modes provides a better, cleaner, higher signal-to-noise probe of reionization.

We explore the impact of reionization topology on 21-cm statistics. Four reionization models are presented which emulate large ionized bubbles around over-dense regions (21CMFAST/ global-inside- out), small ionized bubbles in over-dense regions (local-inside-out), large ionized bubbles around under-dense regions (global-outside-in) and small ionized bubbles around under-dense regions (local-outside-in). We show that first-generation instruments might struggle to distinguish global models using the shape of the power spectrum alone. All instruments considered are capable of breaking this degeneracy with the variance, which is higher in outside-in models. Global models can also be distinguished at small scales from a boost in the power spectrum from a positive correlation between the density and neutral-fraction fields in outside-in models. Negative skewness is found to be unique to inside-out models and we find that pre-SKA instruments could detect this feature in maps smoothed to reduce noise errors. The early, mid and late phases of reionization imprint signatures in the brightness-temperature moments, we examine their model dependence and find pre-SKA instruments capable of exploiting these timing constraints in smoothed maps. The dimensional skewness is introduced and is shown to have stronger signatures of the early and mid-phase timing if the inside-out scenario is correct.

We demonstrate a methodology for optimizing the ability of future dark energy surveys to answer model selection questions, such as `Is acceleration due to a cosmological constant or a dynamical dark energy model?'. Model selection Figures of Merit are defined, exploiting the Bayes factor, and surveys optimized over their design parameter space via a Monte Carlo method. As a specific example we apply our methods to generic multi-fibre baryon acoustic oscillation spectroscopic surveys, comparable to that proposed for SuMIRe PFS, and present implementations based on the Savage-Dickey Density Ratio that are both accurate and practical for use in optimization. It is shown that whilst the optimal surveys using model selection agree with those found using the Dark Energy Task Force (DETF) Figure of Merit, they provide better informed flexibility of survey configuration and an absolute scale for performance; for example, we find survey configurations with close to optimal model selection performance despite their corresponding DETF Figure of Merit being at only 50% of its maximum. This Bayes factor approach allows us to interpret the survey configurations that will be good enough for the task at hand, vital especially when wanting to add extra science goals and in dealing with time restrictions or multiple probes within the same project.