For the most up-to-date listing of my publications, I suggest trying this ADS author query, which gives a comprehensive list including ASCL citations and Atels. Alternatively, see my manually curated ORCiD entry, which is less frequently updated (and limited to more traditional publications), but a little easier to read.
A radio jet from the optical and x-ray bright stellar tidal disruption flare ASASSN-14li
Published in Science (2016)
A high-impact publication about a transient flare event, thought to be caused by a star accreting onto a black hole. The work made extensive use of our software tooling (and close collaboration with the AMI team) to achieve unprecedented temporal-resolution in the radio-lightcurve and a fast turnaround of the analysis.
Coverage in the popular media was quite extensive, examples below:
Chimenea and other tools: Automated imaging of multi-epoch radio-synthesis data with CASA
Published in Astronomy and Computing (2015)
Describes how we combined multiple legacy software components together with Python interfacing code and some novel algorithms to automate a previously labour intensive data analysis task. The outcome is greatly improved response times to transient astronomical events, and perhaps just as importantly, a reproducible reduction process.
The LOFAR Transients Pipeline
Published in Astronomy and Computing (2015)
A thorough reference work laying out the core algorithms behind the TraP, a transients-detection and cataloguing pipeline developed initially for LOFAR. TraP is the first pipeline of its kind to be fully open-sourced and comprehensively documented. Technically, the codebase is interesting for its extensive use of SQL procedures to perform in-database analysis and processing of spatial data. The accompanying data-exploration and visualisation tool Banana is also quite novel in astronomy - we pushed Django to its limits to perform detailed queries on the dataset via a web-based interface that allowed fluid user interaction without complex local installation procedures.
More detail can be found on the code page
Background: A software ecosystem for transient astronomy
Since October 2011 I've been working with the 4 Pi Sky project, which (broadly) aims to automate both the detection and follow-up observation of astronomical transients, motivated in large part by the next generation of large radio telescopes.
Nuts and bolts, transient triggers
A fair portion of my work to date has been developing basic software infrastructure, in the form of various small libraries and tools for automating processing of astronomical alerts and unattended data reduction. This sort of stuff is quite satisfying when done well, but I wouldn't really classify it as 'research' in its own right. However, the efforts are beginning to bear fruit. With the help of our collaborators at the MRAO, since mid-2012 we have been triggering automatic observations with the AMI Large Array, with three papers published to date, and further work in prep. These tools are all open-sourced and publicly available (more details on the code page).
Transient discovery with the 'Trap'
I've also been helping with development of a software project born out of the LOFAR-TKP group at UvA. The Trap is a transients detection pipeline written primarily for use with data from LOFAR - a low frequency radio array. LOFAR will observe at ~10s to 100s of megahertz rather than the usual gigahertz range, giving us a new window on the radio sky. While LOFAR presents a particular challenge due to the sheer volume of data produced, the algorithms being developed are applicable to a range of transient astronomy projects. A paper and open-source release are now available - see the code page for details.
Application of active-learning / Bayesian decision theory to transient follow-up
Finally, the most abstract (but potentially most exciting) aspect of my current research addresses the problem of automating transient follow-up - how to decide what to observe, and when. This is still a work in progress, but I hope to start presenting early results in the next couple of months.
(PhD, 2007 - 2011)
NB link to thesis at end.
My PhD research was focused upon algorithm and software development for a technique called lucky imaging. Normally, ground based astronomy is limited in resolution due to time-evolving localised fluctuations in the refractive index of the atmosphere. Like grease on a camera lens, the atmospheric turbulence blurs the image in a long exposure. However, by taking very short exposures (i.e. at frame rates of around 25 frames per second and above), we can 'freeze' these atmospheric effects and pick out moments when the blurring is at a minimum - so called 'lucky images.' We do this by tracking a bright object in every frame (the 'guide star'), which we also use to align the images before we add them to get the final result. For very bright objects such as planets, you can even do this at home with a small telescope and a webcam.
For many years the application of lucky imaging to general purpose astronomy was limited due to the levels of readout noise at high frame rates. However, with the development of electron-multiplying CCDs, photon-counting performance at high frame rates became a possibility.
Over the course of my PhD I refined the data reduction techniques for lucky imaging, optimising both the method for tracking the guide star and the thresholding algorithm used to perform photon counting. I also developed a high-performance data reduction pipeline capable of reducing data from a multi-CCD mosaic camera at around 200 megabytes a second - lucky imaging with large format CCDs rapidly becomes a serious data handling challenge.
I also did some work on Monte-Carlo simulation of lucky imaging data, both for simulation of combined adaptive-optics + lucky imaging systems, and for development of more sophisticated wide-field lucky imaging algorithms.
If you're interested in the details you can download my thesis via the following links, or via the arxiv: