PhD Projects with Professor Chris Tinney

Our research group focuses on exoplanetary detection using "Doppler Wobble" techniques. We also pursue studies of brown dwarfs - objects somewhat more massive and brighter than exoplanets, but which we can actually see - in order to gain insight into exoplanetary properties. These areas encompass a large range of possible projects - to give a flavour of the sort of work involved, a few of them are spelled out in more detail below. But this is by no means an exhaustive list.

Our focus in supervising PhDs is to train students for a subsequent career as an independent, astronomical researcher. So you'll be travelling to telescopes in Australia and overseas to do observing, analysing data, writing papers, and presenting results at international conferences - and we have funds to support you in doing all that.

There are two main requirements for becoming a UNSW graduate student,

1. • Being accepted in the Graduate Research School, and
   

2. • Accessing funding to support you while undertaking research full time for 3.5 years.
   

Australian students should see the Graduate Research School web pages for details on how to Apply for an Scholarship at UNSW and how to Apply for Admission (deadline around mid-December and late-May).

Prospective students from overseas (in addition to Applying for Admission will have to pay tuition fees. Again there are Scholarships available (see UNSW International Research Scholarships for details - deadlines are a few months earlier than for Australian scholarships).

International student scholarships are very competitive.  You will need to have a four-year undergraduate degree (or undergraduate degree followed by a Masters degree) equivalent to at least First Class Honours level (in the Australian/UK system) - this is roughly equivalent to a GPA in the US system of better than ~3.6 (see this link for more details). It will also be essential to be able to obtain good references from researchers who know you and can recommend that you have the potential to be a research scientist. If you have published papers from Honours or Masters research projects this will greatly strengthen your case.

Your home country may also have scholarships available to support overseas study.

UNSW requires applying students to have discussed possible projects and organised a prospective supervisor in advance of applying for admission, or for any of these awards. So, if the PhD experience outlined above sounds interesting to you, please contact me well in advance of these deadlines, so we can discuss possible projects.

A few worked examples of projects on offer are listed below. These change constantly with time, so please get in touch to discuss additional projects

Measuring Planet Masses and Densities with Veloce

Space-based transit planet searches (like the NASA Kepler and TESS missions) have revolutionised the discovery of planets orbiting other stars. These transit detection have discovered many thousands of  short-period planets orbiting other stars. They also tell us the size of those planets - but what they don’t give us are planet masses. If we know both a planet’s size, and its mass, then we also know its bulk density - which tells us whether its a gas-giant (like Jupiter or Saturn), and ice-giant (like Neptune or Uranus) or a rocky planet (like Venus or Earth).

The only way to get masses for these planets is to make dynamical measurements using radial velocity data like that obtained by Veloce. we now have data over hundreds of nights for a large number of TESS planet candidates. So there is scope for PhD research on the analysis and interpretation of those data - in particular to determine whether any of those planets are in orbits that might make the planets potentially habitable.

Extracting Better Velocities from Veloce

To get Veloce built for its ultra-low price tag (similar instruments on other telescopes have cost between 5 and 10 times more to build), we had to make some innovations in instrument design - like feeding the instrument with an integral-field unit, and compressing the data on the detector.

This presents new and interesting challenges in the data processing. To deal with these challenges we have built a world-first model of the spectrograph based on the physical optics used and grating equations for the two gratings used. These model the performance of the instrument (over time) with just a handful of variable parameters.

As this modelling system is so new, there is lots of scope for research into better ways to model, extract and calibrate Veloce data - leading to higher precision velocities, higher precision planet masses and the ability to measure smaller mass planets.

New exoplanet surveys with Veloce

In addition to measuring masses for planets discovered by the TESS transit mission, we want to observe those systems for longer periods to probe system architectures – i.e. what other non-transiting planets are present around those stars? Longer period surveys of TESS transiting planets will be used to explore system architectures for a large sample of transiting systems with epochs spaced logarithmically to sample periods from 1d to 4yr, providing a large and complete database of system properties for population synthesis modeling.

We also want to carry out a similar-sized survey of non-transiting stars, delivering a calibration sample without transit-pre-selection. This will form the basis of a longer-term survey of the full system architectures of an optimised low-noise host-star sample at the periods needed to characterise Jupiter analogs. Doppler surveys are the best means to probe planets between 2-10 AU – transit sensitivity plummets at long periods, while direct detection can’t probe inside 10AU. Extant Doppler surveys (e.g. AAPS, HARPS, California & Carnegie) provide our best current data on the frequency of long-period planets, but are limited by intrinsic stellar variability and a history of searching more stars for more planets, instead of exhaustively monitoring a defined sample. We will commence a long-term, legacy survey delivering the best data on outer planet statistics and architectures.

Understanding Intrinsic Variability in Doppler Planet Host Stars

The precisions being achieved by Doppler Wobble exoplanet searches like the Veloce team’s surveys and the AAPS are being continually improved. New planet discoveries are being made at lower and lower masses, corresponding to lower and lower Doppler amplitudes, requiring that planet search teams understand the noise behaviour of their target stars in some detail, because it is the Doppler noise produced by those stars (or rather by their surface inhomgeneities and intrinsic oscillations) that are now a limiting factor.

We therefore need to develop better ways in which to parametrise the surface inhomogeneities of stars, preferably using the spectra we obtain from our Doppler search programs.

One area that has been little explored to date is the use of codes like those used in Doppler imaging to tackel the problem of line shape changes in radial velocity planet-host targets. Doppler Imaging codes typically tackle a different problem, which is to use many observations of the spectrum of a star as it rotates, to acquire a tomographic data set that can then be used to deconvolve back to the 'map' of variations on the stellar surface. Basically these codes track small deviations from the overall spectral line shape that the star would have in the absence of any surface variations. As the star rotates these deviations move across each spectral line, and with enough data on enough spectral lines, you can solve the inverse problem and map the stellar surface.

Figure (due to Strassmeier 2006) showing how dark regions on a stellar surface can map to changes in the shape of a spectral line. Of course, such changes can also produce artifical Doppler planet signals! We would like to be able to quantify the extent to which this takes place.

For Doppler wobble experiments we are interested in the extent to which such surface variations can produce artificial overall radial velocity changes. So while the 30-100 observations made of our Doppler host stars could not possibly be used to obtain a tomographic map, they can be used to estimate the extent to which line variations occur, and to quantify their impact on our over Doppler velocities.

There is scope for a PhD student with a strong data processing and computation background to explore the use of these Doppler Imaging techniques as a means of quantifying the impact of surface variations on Doppler planet detections.

Orthogonal Line Profile Decomposition

A stellar surface rotating at a velocity V can be modeled as a projection, from the 2D Doppler velocity surface of a featureless star (multiplied by a brightness map), to an intrinsic stellar Doppler profile I(v) spanning the velocity |v|<V. This intrinsic profile I is then convolved with the spectrograph instrumental profile S, to give an observed profile, P(λ) =S(λ)⊗I(v), containing encrypted information on the asymmetries in I. Unlike in Doppler tomography, the details of the stellar surface brightness map are not important – only the net asymmetry of the intrinsic profile matters. By choosing suitable orthogonal basis functions, Lk, we can define coefficients ak such that P=Σk=0,n ak S⊗Lk(v/V), with the odd Lk being anti-symmetric. Measuring odd coefficients and integrating the odd functions, directly measures spurious Doppler shift.

So one possible PhD project is to implement this methodology on Veloce data (and archival data from other instruments), to develop new line-profile variability metrics that “clean up” that Doppler data. 

This page last updated by Chris Tinney, 18 July 2023