Research
Accretion, compact objects and the time-domain sky.
I explore how matter behaves in the Universe’s most extreme environments. My research centres on the physics of mass transfer and accretion in compact interacting binaries, in particular the properties and evolution of cataclysmic variables and other stellar remnants.
Grounded in time-domain astrophysics and spectroscopic surveys, I compare observations from space-based timing missions (such as Kepler and TESS) to test whether accretion physics is scale-invariant, from young, forming stars to supermassive black holes. I also model gravitational-wave emission from Galactic binaries to establish their properties for future LISA observations, and apply machine learning, deep learning and pattern recognition to classify and interpret large astronomical datasets.
We present an “unreview” of the field of accreting white dwarfs, identifying critical open questions and unresolved physics, such as angular momentum transport, disc winds and magnetic boundary dynamics, to advance accretion theory across all stellar and galactic mass scales.
We report the discovery of a persistent Hα bow-shock nebula around the magnetic, short-period cataclysmic variable 1RXS J052832.5+283824 (RXJ0528+2838), suggesting a powerful, previously unrecognised energy-loss mechanism that could govern magnetic binary evolution.
We estimate the gravitational-wave signals from the Galactic population of cataclysmic variables (CVs) and evaluate their significance for the LISA mission, finding that key systems will be individually resolvable and contribute substantially to the Galactic binary background.
We present a sub-arcsecond cross-match of Gaia DR3 against the IGAPS and UKIDSS surveys. The resulting XGAPS catalogue provides precise multi-band photometry for over 33 million sources, with Random Forests used for astrometric cleaning.
We propose a model in which the magnetic confinement of accretion streams onto accreting magnetic white dwarfs triggers localised thermonuclear runaways, explaining the rapid optical bursts known as micronovae.
We report optical bursts from TV Columbae, EI Ursae Majoris and ASASSN-19bh. These events, dubbed micronovae, release about a millionth of the energy of classical novae and are triggered by thermonuclear runaways in magnetically confined columns.
TESS optical light curves of TW Pictoris show rapid transitions in which the system drops in luminosity by up to a factor of 3.5 in under 30 minutes, analogous to the transitions seen in transitional millisecond pulsars.
Archival optical observations of MV Lyrae show quasi-periodic bursts every 2 hours, revealing an unstable, magnetically gated accretion mode and providing a new way of measuring weak magnetic fields (104–105 G) in white dwarfs.
We study the K2 light curve of a young M-type pre-main-sequence star showing irregular dimmings of up to 65%, and propose causes such as a warped inner-disc edge or transiting cometary-like objects on tilted orbits.
Accreting white dwarfs and young stellar objects display the same linear rms–flux relation, and obey the same quantitative scaling relation, as stellar-mass and supermassive black holes, establishing the size of the accretor as the primary scaling parameter.
Simultaneous Kepler optical and MAXI X-ray timing reveals bimodality and a discrete correlation indicating a 12.5-hour optical lag. Cepstrum analysis confirms echoes consistent with thermal reprocessing in the disc.
A brief overview comparing the aperiodic X-ray variability of X-ray binaries with the optical variability of cataclysmic variables: once their inner-disc timescales are accounted for, both show similar power spectra, rms–flux relations and Fourier-dependent time lags, in the context of the fluctuating accretion disc model.
We study high-order Fourier bispectra of MV Lyr, V1504 Cyg and Cyg X-1, detecting non-zero, positive biphases at high frequencies that indicate time irreversibility consistent with fluctuating accretion disc models.
We apply the fluctuating accretion disc model to Kepler high-frequency timing of MV Lyrae, deducing a geometrically thick inner flow extending to the white dwarf, with optical variability powered by X-ray reprocessing.
Kepler observations of a newly discovered cataclysmic variable reveal deep eclipses every 3.97 hours and a full outburst. Tracking the eclipse depth and out-of-eclipse light through the outburst points to changes in the accretion disc radius and/or its radial temperature profile.
Simultaneous ULTRACAM u′g′r′ light curves of MV Lyr and LU Cam reveal the first Fourier-dependent optical time lags in cataclysmic variables, with redder bands lagging bluer ones by up to ~3 s (MV Lyr) and ~10 s (LU Cam) at low frequencies.
Using 633 days of Kepler data, we model the broad-band noise of MV Lyrae with multiple Lorentzian components, as done for X-ray binaries, and find variability properties strikingly similar to those of X-ray binaries and active galactic nuclei.
We discover a linear rms–flux relation in the Kepler light curve of MV Lyrae, the first in a cataclysmic variable, mirroring those of black hole and neutron star binaries and active galactic nuclei and pointing to a common origin for broad-band variability.
Worldwide optical photometry and MDM time-resolved spectroscopy confirm 1RXS J165443.5-191620 as an intermediate polar, with a white dwarf spin period of 546 seconds and an orbital period of 3.7 hours.
Analysing over 40,000 INTEGRAL/IBIS pointings against accurate X-ray positions, we show that the point source location accuracy has improved significantly since early in the mission, reducing the sky area to search in follow-up programmes by more than half in some cases.
A global study of hard-X-ray-selected intermediate polars and asynchronous polars from INTEGRAL/IBIS, Swift/BAT and Suzaku/HXD, linking their hard X-ray properties to spin and orbital periods and discussing the results in the context of magnetic CV evolution.
Combining simple metrics, supervised neural networks and visual inspection, we build a catalogue of 3552 broad absorption line quasars from 28,421 SDSS DR5 quasars, and show that no single existing metric gives clean, robust classifications.
