Computational Prospects of GPU Correlators

Ian Tretyakov (UofT)

An overview of the current generation of GPU correlators and upcoming technology driving the field forward.

Modelling the Milky Way through Hydrogen Gas

Tyler Foster (Brandon U)

One moderately successful way of mapping the Milky Way’s neutral hydrogen distribution has been by the method of velocimetric deconvolution: calculating the (x,y,z) coordinates of a packet of gas by its observed brightness temperature distribution in sky and velocity coordinates (l,b,v) where v is the line-of-sight velocity. While fraught with uncertainties and debatable assumptions, the method does produces maps that can be made sense of. In this very pictorial and informal discussion, I will present the results of a summer student project to map the very recently released HI4PI Survey; an all-sky HI dataset based on the Effelsberg-Bonn HI Survey and the Parkes GASS Survey. HI4PI is the new standard that replaces the aging Leiden-Argentine-Bonn (LAB) HI survey, the standard since the 1990’s. New maps of the MW will be shown, comprised of individual thickness or z-height slices of the MW disk from -2 kpc to +3.5 kpc height. The positions of major spiral features will be highlighted, along with the results of an intriguing experiment to map the inner solar-circle of the MW, usually not shown because of the familiar distance-ambiguity that exists there.

A fast, versatile, real-time imaging capability for large radio interferometers

Nithyanandan Thyagarajan (NRAO)

Two of the frontier sciences that constitute the key science goals of next-generation radio astronomy include the evolution of gas and large-scale structure in the early Universe, and the characterization of mysterious explosive phenomena in the Universe through time-domain studies. High sensitivity and time-resolution are essential prerequisites to achieve these goals. Modern radio telescopes such as the Hydrogen Epoch of Reionization Array (HERA), Square Kilometre Array (SKA), Long Wavelength Array (LWA), and Murchison Widefield Array (MWA) are adopting an approach of achieving large collecting areas through hundreds to thousands of small antennas. However, traditional radio interferometry architectures face serious challenges to processing data as they rely on correlating data between antennas, and thus the computing cost scales unaffordably as the number of antennas squared (N^2). I will present the E-field Parallel Imaging Correlator (EPIC), which implements a versatile and efficient algorithm for direct imaging. For large compact arrays, the computational cost is expected to scale much more efficiently as N log N. By design, it can operate at the Nyquist speed of incoming digitized data and can produce science-ready images even on sub-millisecond timescales. Thus, it provides an ideal platform for building image-based fast transient search and monitoring systems for Fast Radio Bursts (FRB), millisecond pulsars (MSP), auroral emission from exoplanets, Solar and Jovian bursts, meteor showers and lightning, and other mysterious astrophysical events. EPIC holds significant advantage and promise for most modern radio interferometer arrays such as HERA, LWA stations, CHIME, and cores of MWA and SKA1-low. EPIC is funded by the NSF ATI division to deploy a real-time GPU-version on the LWA telescope in Sevilleta (New Mexico). I will discuss the successful preliminary deployment of EPIC on the LWA with results including high time-resolution, all-sky images obtained in real time. I will also outline the future plan and the potential for expanding the scope and deploying this ultra-efficient science-ready architecture on other modern radio telescopes.

NFIRAOS: The Facility Adaptive Optics System for the Thirty Meter Telescope

David Andersen (DAO)

NFIRAOS (Narrow-Field InfraRed Adaptive Optics System) for TMT (Thirty Meter Telescope) successfully passed its final design review in June 2018. I will describe the science drivers for NFIRAOS and the design we developed to meet these requirements. In particular, I will focus on some of the trade studies we performed and how those choice will impact the science that TMT, NFIRAOS and the client science instrument IRIS hope to achieve.

Structure in the Milky Way

Trey Wenger (University of Virginia)

The structure of the Milky Way is a constraint on theories of Galactic formation and evolution. HII regions, the zones of ionized gas surrounding recently formed high-mass stars, are important tracers of structure in both the Milky Way and other galaxies. I will summarize the four projects of my dissertation, in which we investigate the morphological and chemical structures in the Milky Way disk. Kinematic Distances: We develop a novel Monte Carlo kinematic distance method and compare the kinematic distances and parallax distances of 75 Galactic high mass star forming regions (HMSFRs). The median difference between the Monte Carlo kinematic distances and parallax distances is 17% (0.42 kpc). We find that, for a large portion of the Galaxy, the kinematic distance uncertainty is up to 10 times smaller than the parallax distance uncertainty. Southern HII Region Discovery Survey: The census of Galactic HII regions is vastly incomplete in the southern sky. We use the Australia Telescope Compact Array to measure 4-10 GHz radio continuum and hydrogen radio recombination line (RRL) emission from HII region candidate targets. The survey has discovered 295 heretofore unknown Galactic HII regions, which increases the number of known nebulae in the surveyed zone by 82% to 568. Galactic Metallicity Structure: The metallicity structure of the Milky Way disk stems from the chemodynamical evolutionary history of the Galaxy. We use the National Radio Astronomy Observatory Karl G. Jansky Very Large Array to measure the RRL-to-continuum brightness ratios of 82 Galactic HII regions. We then derive the electron temperatures and metallicities for these nebulae. We find an oxygen abundance gradient across the Milky Way disk with a slope of 0.052^{+0.004}_{-0.003} dex/kpc. We also find azimuthal structure in the metallicity distribution, consistent with simulations of Galactic chemodynamical evolution influenced by spiral arms. Galactic Morphological Structure: Finally, we explore the spiral structure of the Milky Way by constructing a simple morphological model. The model posits the neutral gas, molecular gas, and HMSFR distributions and includes parametrizations of the kinematics of the Galactic disk, the morphology of the spiral arms, and the warp of the Galactic plane. We use a Bayesian Markov Chain Monte Carlo analysis to estimate the optimal model parameters that reproduce the observed Galactic longitude, latitude, and velocity distributions of HI 21 cm hyperfine emission, CO (J=1-0) emission, and the number of Galactic HII regions.

The Large-Scale Structure of Magnetic Fields Associated with Filamentary Molecular Clouds

Mehrnoosh Tahani (University of Calgary)

We present a new method to find the line-of-sight strength and morphology of magnetic fields in star forming regions using Faraday rotation measurements. We applied this method to four relatively nearby filamentary molecular clouds of Orion A, Orion B, Perseus, and California. In this method we use rotation measure data from the literature, a chemical evolution code, along with extinction maps of each cloud. In California and Orion A, we find clear evidence that the magnetic fields at one side of these filamentary structures are pointing towards us and are pointing away from us at the other side. The magnetic field morphologies that can explain this change of direction across filaments include toroidal, helical, and bow morphologies. We investigate these three models by combining our results with those of Planck observations in Orion A. We find that of the three possible morphologies, toroidal is the least probable one and we suggest that the bow morphology is the most natural one. To investigate these morphologies further, we use magnetohydrodynamics simulations to simulate filamentary molecular clouds and their magnetic field evolution. We find that helical fields are not easily generated in our tested scenarios.

The dark side of globular clusters

Vincent Hénault-Brunet (DAO)

The central regions of globular clusters (GCs) have become fertile grounds to search for black holes and understand their formation. The detection of gravitational waves from merging stellar-mass black holes (BHs) by LIGO/Virgo has led to the suggestion that dynamical interactions in the dense cores of GCs may be a prime formation channel for these BH mergers. However, since accreting BHs are rare, we currently do not know if BHs are ubiquitous or not in GCs. Intermediate-mass black holes (IMBHs; ~10^2-10^5 solar masses), possible seeds from which supermassive black holes grew in the early Universe, have also been suggested to lurk in the centre of GCs that survived to the present day. Evidence for ‘leftover’ IMBHs in GCs however remains controversial. In this talk, I will present applications of multi-component models to dynamically infer the presence of a stellar-mass BH population in GCs based on the effect of BHs on the phase-space properties of visible stars. I will also discuss related implications for reported detections of IMBHs, as well as future prospects for confirming those claims or alternatively ruling out the presence of IMBHs in GCs. Moving out to the poorly explored low-density outskirts of GCs, I will discuss how proper motions from the Gaia satellite are enabling efficient spectroscopic follow-up of distant cluster members. This will allow us to disentangle the effect of dark matter halos, Galactic tides, and relics of progenitor dwarf galaxies on the chemistry and dynamics of GC outskirts.