About me
I studied physics from 1973 to 1976 at the University of Manchester in England, and subsequently studied radio astronomy at Jodrell Bank. I received my PhD in 1980 for work on the applications of Bayesian methods to radio astronomical imaging. Immediately following my PhD, Peter Wilkinson and I developed the self-calibration algorithm widely used in radio astronomy. In 1980, I moved to Socorro, New Mexico to work on the newly completed Very Large Array telescope run by the National Radio Astronomy Observatory. Over the 25 years at the NRAO, I was fortunate to work with excellent colleagues on the best radio telescope in the world. In that stimulating environment I was able to make a number of contributions to radio astronomical techniques, including the key algorithms needed for wide fields of view.
At NRAO I led with deconvolution work for the VLA (especially maximum entropy but also other applied mathematical frameworks), work on the theory and algorithmics of mosaicing, triple correlation imaging, the use of supercomputers in astronomy. I like to move into relatively unexplored areas or bring new approaches to old problems. As an example of the former, together with Peter Napier, I wrote an early paper (1988) on using focal plane arrays to correct radio telescope aberrations, in a direct analog of the self-calibration technique used in radio synthesis imaging. I pursued this insight into radio scintillation, confirming by an experiment (with Ramesh Narayan and Anantharamaiah) the effect of scattering on the spatial coherence function of the scattered electric field. As an example of bringing new approaches to old problems, my 1988 work on optimum arrays for radio synthesis bought simulated annealing into the topic, and has subsequently lead to many more sophisticated approaches.
In 1988, my research career was recognised by NRAO with the grant of tenure.
My initial leadership and management experience was at NRAO, where I managed the commissioning of the VLBA for 2 years, led the AIPS++ project for 6 years and then became head of Data Management for 2 years.
The commissioning of the Very Long Baseline Array was in (slow) progress at the time (Dec 1992) when I took prime responsibility. With the help of some careful prioritisation, the diverse team of scientists, engineers and operations staff worked through numerous technical and logistical problems to bring first science from the telescope.
After the VLBA commissioning, I was selected by the AIPS++ Consortium to manage the AIPS++ Project. AIPS++ was an ambitious and very substantial undertaking to develop software in an international consortium on a scale never seen before in radio astronomy (approximately 250 person-years of effort). I took over AIPS++ at a time when it was in dire trouble (some 5 years into its lifetime). I worked to bring sensible and structured management to the project, ensuring its continuance such that it is now the core of the package of choice (as CASA) for ALMA and the Expanded Very Large Array. AIPS++ was radio astronomy’s first attempt at a substantial international software collaboration, and required a change of culture in software development from unmanaged gurus, to managed capable (distributed) teams with a strong focus on milestones and scientific completeness. Mostly due to the difficulty of that change of culture, AIPS++ was only a partial success. However, it did provide the software foundation of a number of important initiatives such as the Parkes Multi-beam project, and the core software library, casacore, used in both CASA and ASKAPSoft.
During my time as the NRAO Assistant Director for AIPS++, I continued significant research innovations, including continuum subtraction, source noise in radio synthesis images, multifrequency synthesis, refractive scintillation in the solar wind, the lack of a radio shell surrounding the Crab Nebula, and the AIPS++ software framework and implementation (with Mark Wieringa of the ATNF) for the Hamaker-Bregman-Sault Measurement Equation.
As NRAO Associate Director for Data Management (from 2001 onwards), my brief was to bring structure to the full life cycle of data management (proposal, data taking, processing, data delivery) at NRAO across all of its telescopes. The current director of NRAO chose to terminate this activity in 2003, and so I did not have the opportunity to bring it to completion. I then choose to return to my scientific position at NRAO.
Following my return to the NRAO Basic Research staff in 2003, I led a small group investigating wide field imaging for the EVLA and the Square Kilometre Array. This collaboration brought forth two key innovations: w-projection - a ten fold speed increase in correcting for the non-coplanar baselines effect (in the process explaining this as being due to differential Fresnel diffraction), and a computationally viable algorithm for the self-calibration of pointing errors. I also co-supervised a PhD student from New Mexico Tech, Urvashi Rau, who worked on Bayesian methods for source fitting, and multi-scale algorithms for multi-frequency synthesis imaging. During this time, I also researched into RFI mitigation, understanding the equivalence between adaptive filtering and beam nulling. Kumar Golap, Rick Perley, Sanjay Bhatnagar and I developed and demonstrated post-correlation RFI excision in the image plane.
Throughout my career, I’ve paid keen attention to the commonalities between imaging in different fields, participating regularly in interdisciplinary scientific meetings. Recent examples include the Kavli Institute workshop in Costa Rica on Real Problems for Imagined Computers, and a UCLA IPAM meeting on Astronomical Imaging. My Science paper on “The Applications of Closure Phase to Astronomical Imaging” (1989) made many in the non-astronomical community aware of the powerful techniques being used in radio-astronomy. As a result of this paper, I became involved in early discussions of large optical synthesis arrays.
During my time at the NRAO, I was involved in a number of external projects. Some examples:
- I was a member of the 1990 NRC Astronomy Decadal Review panel on Interferometry.
- In 1990, I was a member of the Image Processing advisory panel appointed by NASA following the discovery of the focus error in the Hubble Telescope.
- Starting in the mid-nineties, I consulted for the Naval Research Laboratory Remote Sensing Division on a number of projects. One outcome was a memo on interferometric imaging of extended objects in the near field.
- Between 2001 and 2004, I participated in the NSF/Intelligence Community program Approaches to Combat Terrorism, first as lead of the Image Processing panel, and subsequently as a reviewer of grant proposals.
In 2004, I joined the Square Kilometre Array International Engineering Working Group, primarily to contribute towards computing and algorithms needed for the SKA. This resulted in papers on software costs, processing load, and software development. In 2005, I moved to the CSIRO Australia Telescope National Facility (ATNF) to take the lead role in computing for the Australia SKA Pathfinder (ASKAP). Since then I have led the development of ASKAP computing, most particularly in the development of synthesis software and algorithms, and the provision of high performance computing for the telescope.
Since I joined CSIRO, I’ve worked on a wide range of research topics - the radio/infrared correlation, the ATLAS survey, the nearby radio galaxy Centaurus A, and an advanced formalism for calibration and imaging in radio synthesis. In 2009, I was very honoured to be asked to write a commentary on Jan Högbom’s classic and vastly influential paper on CLEAN.