Dissertation projects for the MSc by Research in Astronomy & Astrophysics

Below are some of the MSc(R) projects for this coming year. However, we encourage prospective MSc(R) students to speak with as many of our academic staff, as early as possible, to find out what we do and to see if there are possible MSc(R) projects available - this list is unlikely to be comprehensive and research projects can quickly become out-of-date. Note that supervisors and projects should be decided within the first 1-2 weeks of the first semester and they should inform the MSc(R) course director, Neal Jackson).

A Northern-sky survey at 15 GHz

Supervisors: Keith Grainge, Anna Scaife

Surveys of substantial areas of the sky at high radio frequencies are challenging and few have been performed. As a result our knowledge of the high frequency radio source population is poor. The Arcminute MicroKelvin Imager Small Array (AMI-SA) is now conducting a legacy survey of the complete Northern hemisphere at 15 GHz. This project will calibrate and image the AMI-SA drift scan data and combine the different scans into a continuous map. This map will be searched for sources, a catalogue compiled and a weighted differential source count calculated. Identified sources will be followed up and imaged with the AMI Large Array to give angular resolution comparable to the NVSS survey. Comparing the AMI catalogue with NVSS will allow calculation of the radio spectral index distribution between 1.4 and 15 GHz and identification of any currently undiscovered inverted spectrum sources.

Bayesian Neural Networks for Classification in Astrophysics

Supervisor: Anna Scaife

Standard neural networks optimise point-wise estimates of the weights for each of their neurons. In contrast Bayesian neural networks assign a probability distribution to each weight instead of a single value or point estimate. These probability distributions represent the uncertainty in the weights and are propagated through the network to provide an uncertainty on the final predictions. This has been reported to reduce over-fitting to specific training data sets and to allow such networks to train more general models from fewer data samples. In this project the student will make a comparison of classifications using standard neural networks with those from an equivalent Bayesian neural network, examining the posterior probability distributions in each case. This will be done for specific astronomical classification problems to examine whether the application of Bayesian neural networks can result in more reproducible classification results across multiple test data sets.  This project will require good computing skills and previous experience with Python is desirable.

Molecular gas in ULIRG-to-QSO transition objects

Supervisors: Dr. Robert Beswick & Dr. Jeff Wagg (SKA Organisation)

Since the early surveys of FIR continuum emission in ultraluminous infared galaxies with IRAS, it has been believed that some galaxies go through a heavilly dust obscured starburst phase before emerging as an optically luminous quasar. These have termed the ULIRG-to-QSO transition objects. The initial burst of star-formation is thought to be the product of a gas-rich merger. However, recent years have seen the emergence of a new picture, where gas-poor galaxies containing an active galactic nucleus, can be `refueled' by mergers with gas-rich galaxies. Sensitive interferometers observing at mm-wavelengths allow us to isolate the source of the molecular gas in these systems previously classified as ULIRG-to-QSO transition objects.

We have completed an interferometer survey of CO line emission in a sample of ULIRG-to-QSO transition objects using the Combined Array for Research in Millimeter-wave Astronomy (CARMA). All of these sources benefit from high sensitivity imaging from the Hubble Space Telescope. A student would work on the CARMA data in order to learn the fundamentals of mm-wavelength interferometry and produce maps of the mm-wavelength CO line emission. It is expected that a research publication would be produced as a result of this analysis.

Fundamental Physics Using Spider Pulsars 

Supervisor: Dr. Rene Breton

Binary pulsars are amazing laboratories to study fundamental physics due to their extreme nature (extreme density, magnetic fields and gravitational field), which is possible as they offer proxies via multi-wavelength observations to measure their physical parameters. Of particular interest are the pulsar binaries nicknamed after deadly spiders, in which a rapidly rotating millisecond pulsar is gradually destroying a low-mass companion. Current observations demonstrate that these particular systems harbour some of the fastest spinning and heaviest pulsars known to us. 

In this project, the PhD candidate will join a dynamic team of several researchers focused on the study of ‘spiders’ with the ultimate goal of furthering our fundamental understanding of physics. They will contribute to projects ranging from the analysis of multi-wavelength (optical and radio) observations aimed at detecting new and characterise known spider binaries, to performing theoretical modelling of the evolution of these systems.

We are particularly interested in candidates having strong technical skills as this work will heavily rely on the use of modern data science techniques such as Bayesian modelling, machine learning and high performance computing. 

Heating the solar corona: 3D numerical magnetohydrodynamic simulations and 1D modelling of the turbulent dynamo

Supervisor: Prof. Philippa Browning

A major unsolved problem in astrophysics is to explain why the solar corona is at a temperature of over a million degrees Kelvin (compared with a surface temperature of about 6000 K). Coronal plasma is believed to be heated by dissipation of stored magnetic energy. Much research concerns modelling the dissipation of magnetic energy through the process of magnetic reconnection. There are two potential projects in this area. One project will involve running numerical simulations using the LARE3D code, which numerically solves the equations of magnetohydrodynamics, and analysing the results. This will build on recent work showing that reconnection can be triggered by the onset of the kink instability in twisted coronal loops, exploring a range of magnetic field configurations and exploring the energy release. The second project will use a simplified analytical model of the effects of turbulent reconnection, building on one-dimensional models widely used for laboratory plasmas in which the turbulence is parameterised as a 'dynamo' electric field. These models will be extended to solar coronal loops, requiring some analytical calculations and simple numerical modelling (solution of ordinary differential equations).

High energy particles in solar flares

Supervisor: Prof. Philippa Browning

Solar flares are dramatic releases of stored magnetic energy, emitting electromagnetic radiation from radio to hard X-rays or gamma rays. Large numbers of non-thermal ions and electrons are produced, but the origin of these high-energy particles is not understood. A promising theory is that particles are accelerated by the strong electric fields associated with reconnection of the magnetic fieldlines. This can be studied using a "test particle" approach, in which charged particle motion is calculated in electromagnetic fields representative of reconnection. A project is available to use test particle models to study the generation of high energy particles in solar flares - this will involve adapting existing computer codes, as well as some programming and analytical calculations.

Probing of the Magnetic Field in Star Forming Regions with SKA using CH

Supervisor: Prof. Gary Fuller

Magnetic fields can play an important role in the formation of stars. This project will investigate the potential of SKA> observations of CH as a probe of the magnetic field in star forming regions. CH has two groups of transitions which can be observed with SKA. The lower frequency transitions, which are at a frequency around 700MHz, are potentially very good probes of the magnetic field. However there has been very little observational or modelling work done on these transitions. This project will investigate the nature of these CH transitions and through modelling, and potentially exploratory observations, how SKA1 and eventually SKA2 will be able to use these transitions to study the magnetic field in dense gas in star forming regions.

SiO Maser Movies

Supervisor: Dr. Malcolm Gray

SiO masers form in broken ring patterns around their host stars, which are mostly long-period variables of Mira and semi-regular variable type. These stars are pulsational variables, and the maser zone is disrupted by periodic shock waves that propagate outwards from the photosphere. The maser zone also coincides approximately with the condensation zone of alumina dust that helps convert pulsational motion into a steady outflow at larger radii. Hydrodynamic solutions for the atmosphere, coupled with maser pumping models, can be combined to provide a theoretical counterpart to a time-sequence of VLBI observations, improving our understanding of the establishment of mass-loss. It should be possible to make a movie of the motions of a number of maser objects at intervals of approximately one month over several pulsational periods.

Contaminating Weak Lensing Cosmology with Active Galactic Nuclei

Supervisors: Dr. Ian Harrison and Prof. Michael Brown

Measuring the weak lensing distortion of the shapes of millions of galaxies across the sky is an excellent way to learn about cosmological parameters and probe the different physical models of Dark Energy. Most weak lensing studies (such as the Dark Energy Survey) are conducted with optical telescopes, but recently new work has begun to focus on the prospect of weak lensing with radio telescopes such as e-MERLIN and the Square Kilometre Array (SKA). Doing weak lensing at radio wavelengths has a number of advantages but also a number of new challenges, one of which is the presence of different types of sources of the sample of galaxies used. The shapes of ordinary star-forming galaxies should be (relatively) easy to model with simple profiles with minimal error, but the shapes of Active Galactic Nuclei (AGN) in the radio are expected to be much more complex. The presence of AGN mistakenly included in the weak lensing sample can cause an error known as 'model bias' which directly affects the cosmological results. The aim of this MSc project will be to investigate how bad this model bias will be for different levels of AGN contamination and how it could be avoided.

The student will perform mostly computational work, simulating images of AGN and fitting simple models to their shapes. The student will gain knowledge in the general areas of cosmology, statistics and radio astronomy. The project has a well-defined outline and could easily lead to a published peer-reviewed journal article.

Gravitational lensing

Supervisor: Dr. Neal Jackson

Gravitational lenses are systems in which a background object is multiply imaged by the action of the gravitational field of a foreground galaxy. Gravitational lensing is important because it allows us magnified views of distant objects in the Universe, and also because it allows us to investigate mass distributions in galaxies independent of their light emission. We are currently involved in a number of projects with major radio facilities (e-MERLIN and LOFAR) and planning for future surveys, including Euclid. Accordingly, there are a number of areas in which students could become involved:

- We are conducting LOFAR observations of a number of gravitational lenses to explore the properties of lensing galaxies. Because lenses give us multiple lines of sight through the galaxy, this allows us to deduce the effect of the lensing galaxy on the radio signal that propagates through it.

- We are currently conducting a survey called LBCS which provides calibrators for the analysis of the long baselines of LOFAR, which use the international stations outside the Netherlands. There are opportunities to get involved with the development of interferometric pipelines for the reduction of LOFAR-LB data.

- We have a number of projects involving radio observations of both radio-loud and radio-quiet lenses in order to study both lensing galaxies and the lensed sources (the latter are visible thanks to the magnification of the lensing galaxy).

- We are involved in a science working group of Euclid which is investigating the use of Euclid-discovered gravitational lenses for the study of galaxy evolution. The student will assist with the simulation of the scientific output from such a survey.

The ages of extra-solar planets

Supervisors: Prof. Albert Zijlstra & Dr. Iain McDonald

How has the frequency of planets changed through the Universe's history, and how do individual planetary systems evolve over time? These questions are important in understanding for how long the Galaxy has been able support life. This project provides answers, by determining a homogeneous set of ages for planets beyond our solar system. The Gaia Space Telescope is returning distances from millions of stars, and we have recently published the temperatures and luminosities for about 1.5 million of these. Several hundred of them are known exoplanet hosts. The MSc project will focus on this targetted set, placing these stars more exactly on the Hertzsprung--Russell diagram, and using the latest stellar evolution codes to determine how old each planet-hosting star is. Comparing this to the star-formation history of the solar neighbourhood will help tell us whether planets are a relatively new phenomenon, or have been present since the Galaxy's early formation, and whether unusual system architectures (e.g. hot-Jupiter systems) are more commonly created early or late in a star's evolution.

Observation and modelling of nova explosions

Supervisor: Prof. T.J. O'Brien

Novae are interacting binary stars in which a white dwarf accretes matter from a companion star. A thermonuclear explosion on the surface of the white dwarf ejects the matter, causing a significant brightening across the spectrum from radio waves to gamma rays. Our group and international collaborators are involved with monitoring nova outbursts using a range of telescopes and interpreting the observations in terms of hydrodynamic models of the explosion. A number of projects are possible in this area, including reduction and analysis of radio observations (including from e-MERLIN), hydrodynamic simulations of shocks and modelling of X-ray emission. Outstanding questions relate to the geometry of the ejecta and the ultimate fate of the white dwarf, in particular whether some of these systems eventually end as supernova explosions.

SETI research with the Jodrell Bank radio telescopes. 

Supervisors: Prof. T.J. O'Brien, Prof. M.A. Garrett, Dr. R. Breton (JBCA) & Dr. A. Siemion (Berkeley) 

We only have one example of life in the Universe, here on Earth. Searches are ongoing on Mars and, in the near future, other solar system bodies.  But we now know there are likely to be billions of habitable planets in the Milky Way alone. We may even find spectroscopic evidence of life (bio-signatures) on one of these planets in the next few decades. However, the unambiguous discovery of intelligent life probably requires the detection of techno-signatures e.g. artificial narrow-band radio signals.  There are a number of ongoing SETI projects around the world searching for such signals (e.g. Breakthrough Listen), and next generation telescopes such as the Square Kilometre Array (see this review paper <http://arxiv.org/abs/1412.4867>) will have a major impact on the field.

At Jodrell Bank, SETI observations can include the analysis of (i) raw voltage data streams from single dishes or (ii) cross-correlated multi-band data from the e-MERLIN interferometer array.  Students working on this project will be involved in recording data and analysing it. A full analysis will also require the students to develop software that sorts and filters the data, permitting potential SETI signals to be distinguished from local contributions e.g. man-made interference. New techniques and algorithms will be explored that take into account phenomena such as Doppler drift, e-MERLIN targeted beam-forming or the detection of low-level, broad-band SETI signals, thus expanding the region of search space that can be explored. The work will be undertaken as part of Jodrell Bank’s collaboration with the Breakthrough Listen programme (see https://breakthroughinitiatives.org/initiative/1).  Students should be comfortable with handling large data sets and the development of software for large computer systems. Although we have no idea whether extra-terrestrials exist, let alone whether they are sending us messages, experience tells us that analysis of signals in a new region of parameter space may well bear fruit in other areas of astrophysics.

A superconducting parametric amplifier for the new generation of radio telescopes

Supervisors: Prof. Lucio Piccirillo, Dr. Mark McCulloch, Dr. Simon Melhuish

Parametric amplifiers are well known to have excellent low noise performances suitable for astronomical receivers. Very recently it has been proposed that the kinetic inductance of superconductors can be used as a non linear parameter to realise ultra low noise parametric amplifiers. These amplifiers have the potential to possess a large gain/bandwidth product and possibly even beat the quantum noise limit!

These amplifiers have also important applications in other fields like, for example, quantum computing.

The student will be involved in the theory of superconducting parametric amplifiers as well as in the design, realisation and testing of test amplifiers. During this work the student will learn radio-frequency design, cryogenic design and testing, low temperatures techniques (sub-K).

Direct detection of very high frequency gravitational waves

Supervisor: Prof. Lucio Piccirillo

Graviton to photon conversion is possible in the presence of a strong static magnetic field (inverse Gertsenshtein effect). The photon generated will be coherent with the original graviton. We are interested in exploring the theoretical and technical issues related to the direct detection of gravitons in the GHz to optical frequencies. The potential sources of such high frequency gravitons are at a moment very speculative: Kalutza-Klein 5D gravity and/or the inflationary period in the big bang. A prototype detector has been in operation for several months collecting useful data.

The student will be involved in the data analysis and simulations as well as in the continuation of the design studies for a more sensitive detector.

A miniature dilution refrigerator for astrophysical applications

Supervisor: Prof. Lucio Piccirillo

Bolometers are the preferred detectors of astrophysical radiation in the mm/sub-mm and far infrared. When operating in low background, bolometers need to be cooled to extremely low temperatures 300 mK to 30 mK. Below 100 mK the preferred cooling technology is based on the dilution cooling of mixture of He-3/He-4. Our group is world leading in the development of miniature dilution systems that can be operated in small cryostats suitable for operations at the focal plane of telescopes.

The student will be heavily involved in the design, realisation and testing of a fully tiltable miniature dilution refrigerator.

Development of a novel passive correlator for the next generation of radio astronomy interferometers.

Supervisor: Prof. Lucio Piccirillo

Diffraction of electromagnetic waves limits the angular resolution of radio telescopes. The larger the diameter of the telescope the better the angular resolution. Unfortunately, building larger telescope to achieve better and better resolution is very expensive: it scales roughly with the diameter to the power of 2.8. Radio interferometry allows us to achieve high resolution by combining the amplitude/phase of the electromagnetic waves coming from many small telescope. The combination is achieved by electronic correlators. Correlators are very complex electronic devices: the current limit to the number of antennas that can be correlated is of the order of few tens (the number of signals, or baselines, goes as the square of the number of antennas).

A novel passive correlator will be studied theoretically and especially experimentally. The idea to be tested consists in optically combining the amplitudes of the waves coming from each telescope in an optical "beam combiner". This system can potentially correlate hundreds of telescopes. The student will work with a team that will design, build and test a 3 x 3 quasi optical correlator working at 11 GHz. The prototype will be fielded at the Jodrell Bank Observatory from where observations of some strong radio sources will be performed.

Machine learning for radio astronomy

Supervisor: Anna Scaife

Simulating realistic representations of the radio sky at varying resolutions becomes increasingly challenging as radio interferometers increase in both sensitivity and spatial dynamic range. In this project the student will design and implement a deep generative model for different classes of galaxies in radio surveys, with a view to producing SKA-scale survey simulations. The project will work initially with data from the NVSS and FIRST radio surveys and will aim to produce simulated versions of these surveys using a generative model. Depending on progress, this work can be extended to include the development of a generative adversarial network for classification of sources within next generation radio surveys

Finding the dark gas in the Milky Way.

Supervisor: Dr. Rowan Smith and Prof. Gary Fuller

Despite the fact that the Milky Way is our home galaxy, there is still much that we do not know. In particular much of the gas that makes up our galaxy is "dark", i.e it cannot be observed using conventional methods. For example molecular hydrogen does not emit radiation at the typical temperatures found in the interstellar medium, instead its presence must be deduced from CO emission, but if the molecular gas does not have CO it cannot be observed. Alternatively if atomic hydrogen is optically thick then its true mass may be underestimated. This is especially true at the interface between cold molecular clouds, where stars like our sun form, and their hotter surroundings.

In this project we will perform mock observations of a recent high-resolution simulation of a Milky Way type galaxy. This will involve taking density, temperature and chemical abundance distributions from the numerical simulation and using a radiative transfer code to calculate the predicted emission. We will use the simulated emission maps to investigate where in the Galaxy "dark mass" might be hidden and what the implications might be for the formation of clouds of molecular gas. In particular we will make hydrogen absorption line profiles that can be compared to the THOR (the Hydrogen, OH and Recombination line) survey to test our models of how molecular clouds form. The comparison to this survey, being carried out by the Jansky Very Large Array, should allow the student to get a flavour of both theoretical and observational techniques. Some knowledge of programming would be beneficial for this project but as we will initially be using existing software and routines it is not absolutely essential.

Recombination line ratios in NGC253

Supervisors: Dr. George Bendo

ALMA has observed multiple hydrogen recombination line transitions at gigahertz frequencies within the central starburst in the nearby galaxy NGC 253. In general, these spectral lines are a superior metric for measuring star formation rate, as the spectral line emission can be directly related to the number of short-lived photoionizing stars that are present (unlike infrared or radio continuum emission) and as the spectral line emission is unaffected by dust extinction (as is the case with ultraviolet and optical hydrogen recombination line emission). However, this is dependent upon the assumption that the line emission is primarily spontaneous emission and not stimulated emission. The goal of this project is to use the ratios of the hydrogen recombination lines to test whether the emission is primarily spontaneous emission.

Unveiling the most extreme starbursts in the Universe with deep, multi-wavelength imaging

Supervisors: Dr. Alasdair Thomson, Dr. Rob Beswick

Studies of the extragalactic background light have revealed that up to half of all the star formation that has occurred throughout the history of the Universe took place in “dusty” environments, which are obscured from the view of visible-light telescopes. Interstellar dust grains absorb UV/optical starlight, and re-radiate it in the far-infrared, giving rise to a well-established relationship between the star formation rate and infrared luminosities of star-forming galaxies. The most luminous dusty galaxies (“starbursts”) formed new stars at rates >500x faster than the Milky Way today, and – though rare – are thought to have hosted up to 50% of the star formation occurring at “cosmic noon”, 10bn years ago.

Observations at ~850µm wavelength (the “submillimetre” regime) are uniquely well-placed to detect dusty starbursts, out to the farthest reaches of the Universe – however, the difficulties in tying emission observed with submillimetre telescopes to individual galaxy populations seen at other wavelenghts have long hindered our efforts to understand this important class of star-forming galaxy in detail. The aim of this MSc project will be to employ a newly-developed technique (Chen et al., 2016) to identify the counterparts to a population of submillimetre-selected galaxies (SMGs) recently identified with the James Clerk Maxwell Telescope in Hawaii, and explore their properties using cutting-edge multi-wavelength datasets. Some prior knowledge of programming and data analysis would be beneficial, but neither are essential at the outset, as it is expected that the student would develop these skills during the course of the project.

The polarization properties of radio galaxies

Supervisors: Dr. Lee Whittaker, Prof. Richard Battye and Prof. Michael Brown

Polarization information in the radio observations of distant galaxies could be extremely useful for future high precision cosmology. For example, it has been shown that there is a correlation between the direction of polarization in radio sources and the gradient of the total intensity field, and this property can be exploited to study the distribution of dark matter in the Universe, the nature of dark energy, and also to place tight constraints on a cosmological rotation signal, such as that expected from cosmic birefringence.

At present, there are various quantities that need to be known for these methods to be competitive, such as the strength of the above correlation and the dispersion in the polarization signal. This project aims to study these properties in observations of radio galaxies, thereby providing an opportunity to contribute to future cosmological studies, such as those performed using the Square Kilometre Array (SKA). 

Fundamental Physics Using Spider Pulsars

▲ Up to the top