Spiel

I grew up in New Zealand and went to school and university there. In 2000 I fulfilled the requirements to graduate with a BSc(Hons) (1st class) majoring in mathematical physics at the University of Canterbury. Although honours degrees usually take 4 years, I did mine in 3. Whilst at Canterbury I was awarded with a Freemason scholarship in 1998, 1999, and 2000. In 2000 I was the top student awarded a scholarship by the Grand Royal Arch Chapter of New Zealand, and received an additional award. After my second year of studies in 1999 I was awarded a scholarship by the University for being amongst the top 5 third-year students.

At Canterbury the grades E, D, C-, C, C+, B-, B, B+, A-, A, A+ were awarded. If you assign 'E' through 'A+' as the integers 1-11, my grade point averages in 1998, 1999, and 2000 were 9.8, 10.7, and 9.8 respectively.

In 2002 I started a PhD programme at the Centre for Astrophysics & Supercomputing at Swinburne University of Technology in Melbourne, Australia. I was awarded one of four Chancellor's Research Scholarships for being the top scholarship applicant that year at Swinburne. I also received a postgraduate research scholarship from the CSIRO. Only two of these were awarded to ATNF applicants.

My PhD at Swinburne was 'Pulsar Applications of Baseband Recorders', although its scope is slightly broader than that. A pulsar is the compact remnant that is left over after a massive star explodes in a supernova explosion. They have some extreme properties:

• A typical pulsar has a mass about one and a half that of our Sun.
• A typical pulsar is about 20 kilometres across. The Sun is over a million kilometres across!
• Pulsars a very dense- one teaspoon of pulsar matter weighs over a hundred thousand tonnes!
• Pulsars can spin very fast- the fastest known spins over 700 times per second. Its surface is travelling at about 20% of the speed of light.
• Pulsars have ultra-strong magnetic fields that approach, and possibly exceed, the quantum critical field. Their fields can be 100 million times stronger than the strongest fields that scientists on Earth can produce.
• Strong currents in the magnetosphere of pulsars lead to the emission of radio waves. These are seen to pulse as the pulsar rotates, and can be seen across the galaxy.
• Millisecond pulsars can be timed very accurately. This allows astronomers to map out their motions in space accurately. Timing of pulses from PSR B1257+12 resulted in the discovery of a moon-mass planet around it, and timing of PSR B1913+16 resulted in the first evidence for gravity waves.
• Ultra-bright pulses from PSR B1937+21 have been modelled to be the brightest object in the Universe. The brightest observed pulses from it have inferred brightness temperatures of 5 thousand billion billion billion billion degrees.

These remarkable objects require remarkable observing equipment. My work involves analysing observations of pulsars that were taken using some of the largest radio telescopes in the world:

Dish size is just one factor that determines how well you can observe pulsars. Because pulsars rotate so quickly, their emission changes on very short timescales. Emission from the fastest pulsars need to be observed millions of times per second across wide frequency ranges. Our group at Swinburne University has pioneered a new technology for such observations- wide-bandwith baseband recording.

The CPSR2 baseband recorder was installed at Parkes Observatory in 2002. This instrument samples two 64MHz wide dual-polarisation bands at the Nyquist rate. This equates to measuring the voltage induced in a receiver 512 million times per second, which means sustained recording of data at 128 MB/sec. Data is usually farmed around a cluster of 30 computers for storage or processing.

When CPSR2 was installed it was the widest bandwidth baseband recorder ever built. It has a total bandwidth of 6 times that of its predecessor, and acquired data at a much faster rate than any previous pulsar backend. Despite the challenges of 'pushing the envelope' like this, CPSR2 provided an opportunity to develop software tools to analyse baseband pulsar data in new and exciting ways. This was my thesis- to find new applications for CPSR2 and other baseband recorders.

The initial idea of my thesis was to conduct the world's first pulsar survey that used a baseband recorder. This didn't work out because:

• The high data transfer requirements of CPSR2 really pushed networking technology in a way that wasn't conducive towards sustained data acquisition for pulsar searches.
• There wasn't enough storage for the survey.
• There wasn't enough computing power available for the survey.
• Both CPSR2 and the receiver for the survey came online quite a while after I started my PhD.
• Digital TV came online in New South Wales that interferred severely with observations.

Two years into my PhD programme I started working on using baseband recorders to search for and study ultra-bright pulses from millisecond pulsars at very high time resolution. This research has been highly successful. Here are some highlights:

• My research has lead to the discovery of weak pulses that share all the other characteristics of 'giant pulses'. Therefore, the term 'giant pulse' is a misnomer, and I've introduced the new term 'ultra-bright pulses', which is actually accurate.
• Prior to my research, there were only 2 millisecond pulsars known to emit ultra-bright pulses. I found such pulses from another 3 millisecond pulsars, and so increased the known population by 150%.
• I showed definitively, for the first time, that giant pulses are not manifest in the millisecond pulsar population as a whole.
• I found that PSR J1823-3021A is easier to observe through its ultra-bright pulses than its integrated emission. This is the first millisecond pulsar found to have this property. It implies that a currently unseen population of neutron stars exist that are only detectable through individual ultra-bright pulses.
• I found ultra-bright pulses from the Black Widow pulsar. This emission is so rare, that even with the world's largest fully steerable telescope and microsecond sampling times, only 4 seconds of this emission is visible per millenium !
• I've shown that ultra-bright pulses from PSRs J0218+4232 and J1824-2452A occur in narrow phase windows that align with X-ray components. I've therefore shown that this property is a general characteristic of ultra-bright emission.
• The brightest pulse I observed from PSR J1824-2452A was just 20 ns wide and had an inferred brightness temperature at maximum of 5 x 10^37 K. To my knowledge, this pulse is the brightest object in the universe that has been directly observed and submitted for publication. Publications of observations of PSR B1937+21 will likely change this in the future.
• Observations of PSR J1823-3021A have resulted in the first bound on emission altitude for ultra-bright pulses.
• I have observed the first instance of a double ultra-bright pulse. This implies that the phenomenon results from a global excitation of the pulsar magnetosphere that persists for millisecond timescales.
• I have conducted the first blind searches for recycled neutron stars by searching for individual pulses that have microsecond timescales. This search involved processing dozens of terabytes of data and thousands of hours of CPU time.
• I have successfully carried out some of the most computationally expensive analyses of pulsar data ever. Some pulsar astronomers, such as Matthew Bailes, claim that undertaking a single coherent dedispersion of pulsar data is computationally expensive. For searches for 4 microsecond radio pulses from the X-ray pulsar J0537-6910 coherent dedispersion was repeated 70 times. It takes 60 CPUs to process in real time the data CPSR2 acquires with one coherent dedispersion. And for my searches coherent dedispersion wasn't even the computationally intensive part!
• M15 is the largest globular cluster in our galaxy. Using Arecibo Observatory our group can rule out the existence of pulsars emitting ultra-bright pulses at similar rates to PSR B1937+21 that are stronger than 60uJy at 430MHz. This is very weak - there are expected to be 75 pulsars in the cluster brighter than this. From this result and using the assumption that sub-millisecond pulsars emit ultra-bright pulses at high rates, we have derived a novel limit on the existence on these 'holy grail of pulsar astronomy' objects and can say that they are extremely rare.

The main thrust of my thesis has been developing software tools to use on clusters of computers to allow this science to be achieved. Mostly I've program in C++, although I do know a couple of other languages. I specialise in automating large processing tasks quickly and effectively that involve terabyte-scale datasets. As much as I hate to admit it, studying astronomy actually has qualified me to undertake useful tasks in the real world.