Douglas Scott

We are still trying to understand how galaxies form and evolve. Galaxies bridge the large scales of the expanding Universe and the relatively small scales of stellar systems by tracing the cosmic evolution of matter in individual structures that undergo physical processes. A subset of galaxies, dusty star-forming galaxies (DSFGs), is key to providing insights into the underlying physics. DSFGs emit the bulk of their light at far-infrared to millimetre wavelengths. They are very actively star-forming and are more common in the early Universe, when the details of galaxy evolution remain unclear. With decades of observations, statistical analysis is important to understand the physics behind the whole population of galaxies. In this study, we investigate galaxy number counts, i.e.the number density of galaxies as a function of their flux density, at submillimetre wavelengths using the 'P(D)' fluctuation analysis method. This is a widely used statistical framework that probes the galaxy number counts to the faint end below the detection limit, which is achieved by analyzing the contribution of light from faint galaxies in the overall fluctuations in the map, i.e.the flux histogram. However, P(D) assumes a random spatial distribution of galaxies, whereas galaxies are actually clustered, tracing the cosmic structure. We study the effect of clustering in galaxy number counts using the SIDES (Simulated Infrared Dusty Extragalactic Sky) simulation, which has realistic clustering from the computation of dark matter halos where galaxies reside. We then simulate observed maps for Herschel-SPIRE (Spectral and Photometric Imaging REceiver) at 500?m. We find that clustering biases galaxy number counts. We explore the relation between clustering strength and its bias in the flux histogram, and find that a simple form of correction function can effectively characterize the clustering bias, with its parameters fully determined by the known instrumental properties, together with the two-point correlation function in the map. We test the correction method on simulated maps and improve the resulting galaxy number counts significantly. Our method can be used to revise galaxy number counts in existing data and serve as a powerful tool for future surveys from far-infrared to radio wavelengths.

View record

Estimating the redshifts of distant galaxies is critical for determining their intrinsic properties, as well as for using them as cosmological probes. Measuring redshift spectroscopically is accurate, but expensive in terms of telescope time. Hence it has become common to use “photometric redshifts”, which are fits to photometry taken in a number of filters, using templates of galaxy spectral energy distributions. However, most methods rely on photometry only in the optical and near-IR wavebands, neglecting longer wavelength data. Since the ultimate goal is to obtain redshift estimates for all galaxies, it is important to improve photo-?s for the sources where optical/NIR fits fail to produce reliable results. For specific subsets of galaxies, in particular dusty starbursts, it can be particularly hard to obtain good photo-? estimates, while these same galaxies are often bright at longer wavelengths. Here we describe how to add information from far-IR to millimetre photometry to help improve the photo-? estimates for the dustiest and most actively star-forming galaxies.

View record

Extracting Gaussian information from data is well understood, but characterizing non-Gaussianity is challenging. I describe an approach called the “hierarchical wavelet coefficients” (HWC) method, also known as the “scattering transform”, for analysing full-sky maps and extracting non-Gaussianity information. I introduce a spherical version of the Morlet wavelet and an algorithm using the healpy package to perform the wavelet convolutions. This method is applied to the Sunyaev-Zeldovich and Galactic dust maps constructed from the Planck satellite data to characterize their non-Gaussian features. I propose that in future this method can be used asa test of component-separation methods and robustness of simulations, as well as potentially for cosmological parameter estimation. It can also be used for generating simulated fields with the same statistical features as the real data.

View record

Observations of the cosmic microwave background (CMB) have been fundamental in cosmology for more than 50 years. Many discoveries have been enabled by groundbreaking CMB experiments, including the consolidation of the big bang theory and the determination of the main parameters that describe our Universe. However, there is still more that we can learn from the CMB. In particular, there is the potential to better understand whether the early Universe underwent a period of inflation. To do this we need higher precision measurements of CMB polarization, particularly to detect this so-called ? modes. Measuring ? modes is extremely hard for several different reasons, such as instrumental challenges, foreground contamination and weak signal. B-mode experiments, therefore, must be calibrated extraordinarily well. One possible calibration method is to use known sources in the sky as polarization references. The goal of this study is to provide improved calibration source data, at a wavelength of 850?m. This could help future CMB experiments, including the Japanese-led Lite (Light) satellite for the studies of B-mode polarization and Inflation from cosmic background Radiation Detection (LiteBIRD). We therefore analyse the submillimetre polarization properties of the Crab Nebula, which is the brightest polarized object in the sky at submillimetre wavelengths. We would like to determine its polarization angle to a precision of 0.1° for 850µm light, to fulfill the requirements for CMB ? modes. We use data from the James Clerk Maxwell Telescope (JCMT) and its Submillimeter Common-User Bolometer Array 2 (SCUBA-2) 10,240-pixel bolometer camera. We discuss the details of the data reduction using JCMT's software Starlink, and reach the final value of 141(±16)°, in equatorial coordinates, for the polarization angle of the Crab Nebula. We see large variations in the polarization angle between subsets of the data, indicating that the measurement is dominated by systematic uncertainties (rather than statistical). Although we recognize that there may still be improvements to be made in the reduction process, our conclusion is that it will be extremely challenging to achieve the desired precision.

View record

Submillimetre galaxies have become essential tools in studying the high redshift Universe. Reaching luminosities well over 10¹³ L⊙, they constitute the vast majority of star formation during this early epoch. Their combined infrared and submillimetre emission output is comparable in energy density to all of the optical and ultraviolet light emitted by all of the galaxies in the observable Universe.We have used the Submillimeter Array at 860 μm to observe the brightest submillimetre sources in 4 deg² of the Cosmology Legacy Survey. Previous interferometric studies have found a significant amount of multiplicity at the bright end of the single-dish number counts, suggesting a steepening in the drop-off brighter than 10 mJy, but these studies suffered from small-number statistics. We have targeted 75 of the brightest flux density-ordered single-dish SCUBA-2 sources down to approximately 10 mJy, achieving an average synthesized beam size of 2.4 arcsec and an average depth of 1.5 mJy in our primary beam-corrected maps, corresponding to 4σ detections of about 6 mJy. Our data is sufficient to distinguish between intrinsically bright galaxies and systems that break up into two ≳6 mJy galaxies with flux ratios less than 2 and separated by about 2 arcsec or more, corresponding to a physical distance of around 20 kpc at z=2. We include in our study 28 archival observations of similar nature, bringing our sample size to 103. We statistically deboost our flux density measurements and use these to compute the cumulative and differential number counts of our sample, finding them to be consistent with previous single-dish survey number counts within the uncertainties but with a systematic offset between 2 and 20 per cent. We compute the probability that a >10 mJy single-dish submm source resolves into two or more galaxies with a brightest to second-brightest flux density ratio less than 2 to be about 15 per cent. Assuming the remaining 85 per cent of the targets are ultra-luminous galaxies between redshifts 2 and 3, we find the surface density of >500 M⊙ yr-¹ sources to be 8^{+2}_{-1} deg-² and a likely volume density of 660^{+140}_{-120} Gpc-³.

View record

We examine the brightness of the Cosmic Radio Background (CRB) by comparing the contribution from individual source counts to absolute measurements. We use a compilation of radio counts to estimate the contribution of detected sources to the CRB in several different frequency bands. Using a Monte Carlo Markov Chain technique, we estimate the brightness values and uncertainties, paying attention to various sources of systematic error. At n = 150MHz, 325MHz, 408MHz, 610MHz, 1:4GHz, 4:8GHz, and 8:4GHz our calculated contributions to the backgroundsky temperature are 18, 2.8, 1.6, 0.71, 0.11, 0.0032, 0.0059 K, respectively.We then compare our results to absolute measurements from the ARCADE 2 experiment. If the ARCADE 2 measurements are correct and come from sources, then there must be an additional population of radio galaxies, fainter than where current data are probing. More specifically, the Euclidean-normalized counts at 1.4 GHz have to have an additional bump below about 10 μJy. We present preliminary results of investigating this new population by use of signal stacking. By stacking onto a very deep 1:4GHz radio map at source positions determined in the infrared and optical we hope to be able to see evidence of this population that would be too faint to be seen individually. Results are currently inconclusive. Future work will consist of modelling radio luminosity functions and new observations with the Extended VLA to continue to search for what may be causing this excess emission.

View record

The variability of fundamental physical constants has been a topic of interest boththeoretically and experimentally for many years. Although it is interesting to investigate the consequences of such a variation, it is important to realise that onlythe variation of dimensionless combination of constants can be meaningfully measured and discussed. In this thesis, I try to justify this way of thinking and applyit to two basic cosmological observables, Big Bang Nucleosynthesis and CosmicMicrowave Background anisotropies. I will mention some related studies that areeither wrong or not complete because of being dimensionful.Variation of constants could be considered on two different levels. On the firstlevel one assumes that the constants are time invariant but they can assume differentvalues in different Universes or patches of sky. A thought experiment describinga discussion with aliens having a different system of units with different couplingconstants could be helpful, this idea will be returned to throughout the thesis. Onthe next level, the constants can be promoted to being smooth functions of time orspace. It is good to have a firm understanding of what happens on the previous levelbefore trying to consider genuinely variable constants. For variable constants weneed to consider theories beyond the currently accepted ones, which are capable ofconsistently describing such a variation.I briefly review the scalar-tensor theory of gravity as a possible way to describethe variation of the gravitational coupling. I give a brief historical review on thesubject and consider the theory in the two so called ‘frames’, discussing about thebenefits of each frame mathematically and the physical meaning of these frames.Such theories could form the frame work in which further study of variable con-stants could be carried out.

View record

We have carried out a pilot study for the SCUBA-2 'All-Sky' Survey, SASSy, a wide and shallow mapping project at 850 μm, designed to find rare objects, both Galactic and extragalactic. Two distinct sets of exploratory observations were undertaken, and used to test the SASSy approach and data reduction pipeline.  The first was a 0.5° x 0.5° map around the nearby galaxy NGC 2559.  The galaxy was easily detected at 156 mJy, but no other convincing sources are present in the map.  Comparison with other galaxies with similar wavelength coverage indicates that NGC 2559 has relatively warm dust.  The second observations cover 1 deg² around the W5-E Hɪɪ region. As well as diffuse structure in the map, a filtering approach was able to extract 27 compact sources with signal-to-noise greater than 6.  By matching with data at other wavelengths we can see that the SCUBA-2 data can be used to discriminate the colder cores.  Together these observations show that the SASSy project will be able to meet its original goals of detecting new bright sources which will be ideal for follow-up observations with other facilities.We have also carried out a study of MS 0451-03, a massive galaxy cluster at z=0.5, strongly lensing a group of galaxies at z=3.  Imaging with SCUBA, and more recently with SCUBA-2, shows a prominent arc of submm emission, but lacks the resolution to break up this lensed structure into individual sources.  ALMA, even in Early Science mode, has the ability to finally resolve the giant submm arc.  A lensing reconstruction will allow us to relate the submm sources to lensed objects detected in other wavebands.  The cluster-scale lensing and the extent of the galaxy group make this a relatively simple system to investigate. Lessons learned here could help us understand the effects of differential lensing on the spectral energy distribution in more complicated galaxy-scale lens systems.

View record