Measuring the Metre-wave Sky Spectrum with Milli-kelvin Accuracy: Engineering Challenges and Australian Opportunities


We describe a radio experiment that will test theories of structure formation in the early universe. The 1.42 GHz spectral line of neutral hydrogen, redshifted down to the 45-200 MHz band, traces the ionisation state of the intergalactic medium through the first billion years of the now 14 billion year old universe. Measuring the metre-wave sky spectrum with milli-Kelvin accuracy may thus detect the epoch when the first sources of ionising radiation formed, termed “the epoch of reionisation”. We discuss the engineering challenges presented by this experiment and the opportunities provided by radio-quiet sites in Australia. INTRODUCTION Detecting the epoch of reionisation poses several engineering challenges. We describe a pathfinder experiment over 100-200 MHz which requires a receiving system with an 80 dB dynamic range, 8,192 spectral channels, and a 0.1 dB noise figure. The antenna attached to this receiver must have a pattern which is invariant over the 100-200 MHz band and a very good VSWR. Greater than the challenge of assembling this equipment is the challenge of calibrating it. The bandpass, or relative channel to channel gain, must be calibrated to within 0.00003 dB (1 part in 150,000). These challenges may seem formidable, but they are not as formidable as the contribution to science made by a successful detection. SCIENCE BACKGROUND The best model for structure formation in the universe is the gravitational growth of density perturbations within a hot big bang cosmology. Several independent observations validate this theory, particularly observations of anisotropies in the cosmic microwave background (CMB) and the power spectrum of the spatial distribution of galaxies. We wish to shed further observational light on the dark path leading from the tiny density fluctuations observed in the primordial universe to the large galaxies and galaxy clusters observed in the today’s universe. A challenge for the current structure formation model is to explain the thermal evolution of the underlying intergalactic gas, the intergalactic medium (IGM), as it is bombarded by ionising radiation from the first stars and galaxies. It is difficult to understand the feedback mechanisms that govern the thermal evolution of the IGM because current observations do not provide adequate constraints on the variety of models proposed. Observations of the recent temperature of the IGM, together with measurements of the CMB and the absorption spectra of Quasars, suggest that reionisation must have been a complex, multi-step process [1] [2]. Conversely, simulations based on the current model of early structure formation suggest a simple, single-step process [3]. We aim to constrain structure formation models by providing a measurement of the evolution of the IGM over the first billion years of structure formation. The IGM consists mostly of hydrogen and helium, but we focus specifically on hydrogen due to the prominence of its 1.42 GHz spectral line. We wish to measure this spectral line, averaged over the sky, as it varies with look-back time (redshift) and hence frequency. From this average spectrum we may deduce the ionisation history of the hydrogen in the IGM. IONISATION HISTORY OF HYDROGEN IN THE EARLY UNIVERSE – WHAT WE KNOW Fourteen billion years ago, near the very beginning, all matter in the universe was ionised due to the extremely high temperatures associated with the big bang. About 500,000 years later (z ~ 1088), this plasma cooled sufficiently to recombine and to form neutral hydrogen and helium. This neutral state persisted for at least a hundred million years (to z ~ 30) as implied by WMAP’s measurement of the post-reionisation optical depth to electron scattering [2]. Thirteen billion years ago (z ~ 6), ionising (Lyman continuum) photons began travelling toward Earth’s telescopes from the most distant quasars observed. The spectra of these quasars indicate that the hydrogen in the IGM was predominantly ionised by that time and has remained so until now [4] [6]. The intervening billion years (30 < z < 6), between these two observational points, interests us most. During this period there must be at least one change, of order unity, in the ionised fraction of hydrogen. This implies that there should be a step, of order 10 mK, in the integrated, all-sky spectrum of neutral hydrogen [5] in the range 45 – 200 MHz. A detection of one or more such steps will indicate more precisely when in the first billion years of the universe’s existence the first sources of ionising radiation formed and what they may have been. ENGINEERING REQUIREMENTS Sensitivity Sharp changes in the ionisation state of the hydrogen gas imply sharp features in the background spectrum at the level of a few mK. The level of this signal alone does not challenge us. We assume that the Galactic synchrotron emission dominates the system temperature and follows a simple model of 150 K at 150 MHz with a spectral index of −2.6. This is consistent with cooler parts of the sky at higher Galactic latitudes. Fig. 1 shows that the required 1 mK sensitivity is reached in just 1.4 hrs at the upper end of the band of interest with a 1 MHz channel. If the Galaxy dominates the system temperature, we need not over-optimise the noise figure of the system. We can allow a receiver with an equivalent noise temperature anywhere up to 10% of the Galactic brightness temperature. Fig. 2 plots the resulting noise figure requirement. A receiver with a noise figure of 0.1 dB will be suitable. Integration Time for 1 mK Sensitivity 1 10 100 1,000 10,000 0 50 100 150 200 250 Frequency (MHz) T im e (h rs ) Fig. 1. Integration time to reach 1 mK r.m.s. sensitivity in a 1 MHz channel.


3 Figures and Tables

Download Full PDF Version (Non-Commercial Use)