MSc projects
The projects below are those already proposed specifically for an MSc by Research. However, other projects can be designed and agreed through discussion with staff at the interview stage. Prospective students are encouraged to also look at the PhD project webpages for the types of projects that could be adapted for an MSc and should feel free to contact potential supervisors to discuss options. Each project listed below consists of a title, names of the supervisory team (with the principal supervisor in bold text), and a short description of the project.
-
Predicting habitability across the Universe
Supervisory Team: Rob Yates, Team Member 2
One of the most intriguing questions in astrophysics is "when and where does life form in the Universe?". The answer depends sensitively on many factors, including the chemical composition of the planet-forming gas, the star-formation rate, and the presence of any nearby destructive events such as supernovae (SNe), gamma-ray bursts (GRBs), and accreting supermassive black holes (SMBHs). In this project, you will harness the power of the L-GALAXIES simulation to predict habitability in regions within galaxies across all space and time. You will utilise the simulation output data, which contains millions of galaxies (including their oxygen and carbon abundances, star-formation rates, dust masses, SN+GRB rates, and more) to discover which are best suited to hosting life in the Universe. You will also be able to design new equations for linking galaxy properties to the habitable planet count. This will give us the best picture yet of habitability throughout the Universe from a theoretical perspective.
This project is computational in nature. Therefore, some prior knowledge of python (or a similar programming language) is required.
-
Population level studies of substellar atmospheres
Supervisory Team: Ben Burningham, Fei Wang
NASA’s SphereX mission is providing low-resolution spectroscopic data for 1000s of cool brown dwarfs and isolated planetary mass objects, providing the first population level data spanning 0.6 - 5 microns. In preparation for this, we have developed a neural-network forward model emulator to speed up the process of atmospheric characterisation with retrievals and enable the study of this large data set using techniques previously only possible for handfuls of objects.
In this project, the student will optimise the “Brewster” emulator for studying specific subsets of the SphereX substellar sample, benchmarking the results against the local stellar population and investigating the drivers of the diversity of observed properties of substellar atmospheres.
-
Is Titan's Haze Really Ubiquitous? Testing Atmospheric Models with High-Resolution Methane Spectroscopy
Supervisory Team: Hugh Jones, Martha Mak (external)
Titan is often regarded as the archetype of a haze-dominated planetary atmosphere. However, recent analyses of high-resolution optical spectra have identified numerous methane absorption features (Sirinrat+ 2025) whose strengths and profiles may be inconsistent with current atmospheric models that assume a dominant contribution from haze opacity (Mak+ 2025).
The project will investigate this issue by examining methane absorption features in high-resolution spectra of Titan and Jupiter, including unpublished observational datasets. The student will verify proposed methane line identifications, compare observed spectra with state-of-the-art atmospheric models, and explore how variations in methane abundance, scattering, and haze properties affect the predicted spectra. The study will assess whether modifications to existing models can provide a better representation of the observations.
This work will improve our understanding of methane-rich planetary atmospheres and provide a foundation for predicting the detectability of methane in directly imaged exoplanets. The results will be particularly relevant to future observations with the next generation of Extremely Large Telescopes and their high-resolution optical spectrographs, where methane will be an important biosignature and atmospheric diagnostic.
-
The enigmatic XTE J1710-281
Supervisory Team: Klaas Wiersema, Dom Walton
A low-mass X-ray binary (LMXB) is a binary star system where a compact object (a neutron star or a stellar-mass black hole) accretes matter from a companion star that has a mass less than the Sun. One such example is the neutron star LMXB XTE J1710-281, which has a binary period of 3.3 hours. This LMXB is special because it shows sharp eclipses (where the X-ray emitting part of the accretion disk is covered by the secondary star every orbit) as well as “dipping”: broad dips in X-ray flux that happen once per orbital period when our line of sight to the compact object passes through a thickened region in the outer edge of the accretion disk. This combination of eclipsing and dipping is rare, as it is only possible for a narrow range of inclination angles. The X-ray eclipses of this LMXB have been observed for many years, and some authors have claimed to see unusual changes in the period, and attributed that to a third body in the system, though this is disputed.
Some years ago, we obtained optical and near-infrared spectroscopy (using the X-Shooter instrument on the VLT) of the proposed optical counterpart of XTE J1710-281. In this project, the student will reduce and calibrate these spectroscopic data, confirm (or reject!) the association with the X-ray binary, and investigate the properties of the continuum and lines in the spectra in terms of LMXB accretion disks and secondar stars. This will lead to a better understanding of the accretion disk and secondary star of this enigmatic object.
-
The properties of no-ID radio sources in the LOFAR Two metre Sky Survey
Supervisory Team: Dan Smith, and Team Member 2
The LOFAR Two-metre Sky Survey has produced the best maps of the sky at 150 MHz, detecting millions of star forming galaxies and accreting supermassive black holes. We know this because we have compared the radio (150MHz) properties with data at other wavelengths which enable us to estimate their redshifts (i.e. their distance) and so characterise their properties (e.g. stellar mass, star formation rate, accretion rate). However, there exists a small fraction of LOFAR sources for which we have been unable to identify a counterpart; as a result, their properties are completely unknown. In this project we will tackle that problem – and fill in a major gap in our understanding of the low frequency Universe. This is an exciting opportunity to generate new knowledge – and is great preparation for a research degree in future.
-
Changing-Look AGN in new Spectroscopic Surveys
Supervisory Team: Dan Smith, and Team Member 2
With new spectroscopic surveys we have repeat observations of large numbers of extragalactic objects, taken months to years apart. This offers the chance to identify active galactic nuclei (AGN) that have spectra different from how they looked in the past. There are two types: (i) changing obscuration AGN, where dust has moved along the line of sight, but the physics of the source is unchanged, and (ii) changing state AGN, where something more fundamental has changed to the accretion process, and this manifests in the spectrum by changes to the emission lines. In this project we will search for changing-look AGN, and attempt to quantify how common they are, and on what timescales they change. We may also be able to study the properties of their host galaxies, and how they relate to those of the broader AGN population. This project is an exciting opportunity to generate new knowledge – and it is great prepation for a future PhD.
-
Finding the weirdest radio sources in the Universe
Supervisory Team: Dan Smith, Luke Holden, Soumyadeep Das
The WEAVE-LOFAR Survey is producing more than a million optical spectra of sources identified in the 150 MHz maps from the LOFAR two metre Sky Survey (LoTSS). This is orders of magnitude larger than any previous study of the optical properties of radio sources, and therefore the unexpected is virtually guaranteed. In this project, the student will inspect the WEAVE-LOFAR data set using a combination of visual inspection and automatic methods (e.g. Outlier detection using a random forest algorithm, MCLOF, Self-organised maps, Chi2) to identify spectra of objects that do not fit in to the traditional classes of source in the faint radio source population, and determine their nature.