Hooked on Astronomy
As told by Andrew Fabian
I was hooked on astronomy by the age of about seven, having read in a children’s encyclopedia that astronomers could work out the composition of a star from the light it emits. That seemed wonderful. I recall seeing Comet Arend-Roland when I was nine and I followed the emerging Space Progammes and spent time studying the night sky with a one-inch refractor from our back garden. Asthma meant that I missed junior school several days a month but, provided I sat still, I could read which was the main way my horizons expanded. We had no television till I was about nine. My parents were shopkeepers and not interested in science but left me to experiment, learn the constellations and read by myself.
"At home in the evenings I did electronics with crystal sets, thermionic valves and then transistors, which were just becoming affordable."
After the village junior school, I went to the state grammar school in nearby Daventry where I enjoyed most lessons but especially physics and chemistry. My asthma soon disappeared. At home in the evenings I did electronics with crystal sets, thermionic valves and then transistors, which were just becoming affordable. At 15, I ground and silvered a six-inch mirror and assembled a simple Newtonian telescope. The Moon at 200 times magnification was magnificent. Space and astronomy drew me in and I resolved to take study physics. At 17 I spent a couple of weeks at the Jeremiah Horrocks Observatory in Preston, Lancashire, to gain some experience. Much of my time was taken up by measuring and counting sunspots on drawings that had regularly been made a few decades earlier. It was far from exciting but did not dissuade me from wanting to find out more about the Universe.
I studied physics for my first degree at King’s College, London. Having lived in a village until then, I felt ready for a city and London seemed right. It was an exciting place at times in the late 1960s, although for a penniless student keen on studying science, opportunities were limited. Astronomy did not feature in my lecture course, although I do clearly remember Professor Herman Bondi giving a lecture on “Why is the Sky Dark at Night?” No visual aids, no black board, just clear speaking – taking something which sounds too obvious to discuss and extracting profound cosmological consequences!
For my PhD I considered several options, including radio astronomy at Cambridge led by Martin Ryle and space astronomy at University College London with Robert Boyd. I chose the latter and started my research in the autumn of 1969 at the Mullard Space Science Laboratories in the Surrey Hills between Guildford and Dorking. By December, I had changed supervisors several times and my new supervisor, Pete Sanford, suggested I write a proposal for a Skylark sounding rocket to observe the granularity of the X-ray Background. He had been at a conference that summer where Martin Rees had discussed the origin of this background radiation in terms of seven radio galaxies per square degree. If true, then the granularity should be measurable. I travelled to Cambridge to meet Martin and was deeply impressed by his friendliness and the generosity with his time to someone who was just starting out. I also consulted David Lindley of the UCL statistics department about how to obtain limits and was told to read his books. A proposal was submitted before Christmas and accepted in January. Things could happen rapidly.
The proportional counter detector was the workhorse of X-ray astronomy back then. The X-ray Background was going to be readily detectable, but what I needed to do was reduce the non-cosmic background in the detector that was due to cosmic rays. Pete Sanford had devised a pulse-shape discrimination method for doing that (X-rays produce a compact cloud of electrons in the detector whereas cosmic rays leave an extended cloud). My immediate task was to design the electronics, using integrated circuits which had not been used for that purpose before at the Mullard Space Science Laboratories. I gave myself a crash course in electronics which was far removed from my home lab work years before. After some months it came together, and by autumn I was testing the assembled equipment on a Skylark payload module.
Weeks were spent trying to suppress radio frequency interference. A transmitter was only a few metres away from the very sensitive preamplifier which detected minute electrical signals from the detector. All sharing the same powerlines. Sometimes I would think it was working well, then step back and everything would go haywire. Eventually it was suppressed and the instrument became robust. At the same time I was learning about X-ray astronomy and astronomy in general. At the time it was reckoned that the total exposure to the X-ray sky by rocket-borne detectors was just a few hours, meaning that I could easily read and digest every paper written on the subject in my spare time.
Skylark SL1001 was launched from Woomera in Australia in late January 1971. I spent six weeks in Australia having flown there on a three-day, Ministry of Defence, turbojet flight to Adelaide followed by train to Woomera out in the desert. (The name Woomera is the indigenous name for a throwing stick.) The flight gave about 15 minutes of exposure to cosmic X-rays during the upper part of its trajectory. Fortunately the data were telemetred down during the flight as the parachutes became tangled and the payload smashed to pieces on hitting the ground. I spent the next day in a helicopter making the recovery, which was exciting at first but, as it was extremely hot outside and the desert was dotted with salt pans, there were strong convection currents: we went up and down like a lift, and I spent the last hours feeling nauseous.
I obtained the telemetry tapes some weeks after my return and read them onto an IBM mainframe in London, analyzing them at the space laboratory. The results became my first publication, “Rocket Observations of the Cosmic X-ray Background” by Fabian & Sanford, Nature Physical Sciences, May 1971. Publication could happen fast back then. I felt that I was in competition with the X-ray astronomy group of Riccardo Giacconi at American Science and Engineering in the US. (Riccardo started cosmic X-ray astronomy with a US rocket flight in 1962.) The American Science and Engineering group had launched a satellite from Kenya named Uhuru which had several proportional counters back in December 1970. By my launch they had buckets of data. Fortunately for me they were making new discoveries every week (X-ray binaries, X-rays from supernova remnants, clusters of galaxies and so on) and didn’t get around to the X-ray background. As luck had it, I was generously given and published the Uhuru background data after a visit to them in 1975.
The rocket results showed that the background was very smooth, requiring more than two million sources over the whole sky, and was to be confirmed and extended with a second rocket flight, this time from the European Space Research Organisation (later becoming ESA). Preparation for that meant many trips to the European Space Research and Technology Centre in Noordwijk in the Netherlands, and the launch successfully took place from Sardinia, Italy, in June 1972. In the meantime, I had studied the problem of fluctuations in a background of point sources and found that it overlapped with observations of radio sources in what was known as P(D) – the probability distribution of deflections in the pen recorder as a radio telescope scanned the sky. Through Martin Rees, I was introduced to Peter Scheuer of the Cambridge Radio Astronomy Group who had studied the problem 15 years earlier, and also Dennis Sciama, then at Oxford, who had a student looking at it. Both were interested, friendly and helpful.
A couple of years later I worked on the origin of the X-ray Background with cosmologist Michael Rowan-Robinson and later in the 1980s, Xavier Barcons, with whom I wrote a review on the topic in 1992. In a way we were finding why the X-ray Sky is dark at night. Riccardo Giacconi and others finally resolved most of the X-ray Background into distant active galactic nuclei with the Chandra Observatory in the early 2000s.
Continuous work on X-ray data
My PhD viva was in July 1972 and rather rushed as I thought I was about to head off to Cambridge Massachusetts to work at American Science and Engineering with Giacconi’s group. The problem was the visa. They were then hard to get, and I waited and waited until February 1973 and learned that the job no longer existed as the High Energy Astronomy Observatory project I was to be funded on The AXAF Science Working Group where I am second from left. Nobelist Riccardo Giacconi (who started X-ray astronomy) is 8th from left was cancelled by NASA. It was reinstated a few years later, but I was no longer going to the US. I was lucky to remain at Mullard Space Science Laboratories as a postdoc and began working on the small X-ray detectors just launched on the Copernicus satellite. Pete Sanford was the Project Scientist for the instrument and indeed had spent most of his time in the US over the previous two years. I was probably the only person working full time on the X-ray data for the first six months or so. The satellite carried a UV observatory taking spectra of bright O stars. We could choose the pointing direction for about 10 percent of the time. We looked at X-ray binaries, supernova remnants, clusters of galaxies and active galaxies, which led to many discoveries and much reading, studying and understanding diverse processes. I learned an enormous amount of astronomy and astrophysics. One key target we observed was the Perseus cluster of galaxies, showing that its X-ray emission peaked around the central galaxy NGC1275.
I successfully applied for an Semiconductor Research Corp postdoctoral fellowship to work with Martin Rees who was then a professor at Sussex University. A few months later Martin had accepted the Plumian Professorship at Cambridge, so I joined the Institute of Astronomy at Cambridge in October 1973. That summer there was a conference on compact objects that I attended and heard talks from many of the leading theorists on neutron stars and black holes.
"I was hooked even more."
New questions and opportunities
Over the next eight years I held several postdoc positions at the Institute of Astronomy including the first UK Radcliffe five-year Fellowship in Astronomy. I was using whatever X-ray data I could obtain and also tried theory, finding my math skills were not up to easily solving differential equations and my computer programming was not up to the precise standards necessary for detailed numerical work. I was coming up with lots of questions and ideas and beginning to work with bright research students and my own postdocs on their solution, both observationally and theoretically. I enjoyed the phenomenology of astrophysics.
Around 1980, I was tempted by an offer of a Professorship at Utrecht University, which though attractive would have been a major upheaval. Initially it was at a junior level but that changed in 1981. However, by that autumn I was offered and accepted a Royal Society Research Professorship starting in 1982 and held at the Institute of Astronomy. I retained that post until 2013 and can say it was a great privilege and a dream job. It provided for my professorial salary, research expenses and often a postdoc. I was not obliged to teach but did lecture to final year Physics students on Relativistic Astrophysics initially, and later combined with Cosmology from Anthony Lasenby for the past 20 years. Owing to arcane rules the university classified me until 2003 as an “unestablished research worker,” which is not quite as bad as it sounds. The main thing was that the Institute of Astronomy supported me and my growing research group of students, postdocs and visitors.
In 1983, I became a Fellow of Darwin College, where I could happily chat to physicists, chemists, biologists and others from the Social Sciences, Arts and Humanities. My delight in exploring everything found an outlet in the annual Lecture Series, starting in 1986 with Origins. I have in total co-organised six of the series. The Lecture Series has become the largest public series in Cambridge and are still going strong; I was recently part of the discussion preparing for 2022. I also did a stint as Vice-Master of Darwin for 14 years.
In 1977, my student Paul Nulsen and I explored the consequences of radiative cooling in the cores of clusters of galaxies, prompted by the Copernicus observation of the Perseus cluster and similar hints from other clusters and workers. This lead to our paper on cooling flows published after similar work by Len Cowie and James Binney. We related the expected cooling flow to the enormous optical H alpha nebulosity seen around NGC1275, which must surely be connected. Under the generosity of Giacconi’s group, I visited the Center for Astrophysics in the other Cambridge for two months in 1979 to work on data from the recently launched Einstein Observatory (the third mission of the revived High Energy Astronomy Observatory program). It was wonderful to see and work on many images from the powerful X-ray telescope that it carried. This included the Perseus cluster, the images of which confirmed and extended the work on cooling flows.
President of the Royal Astronomical Society
Around 1983, I joined the Science Working Group of NASA’s Advanced X-ray Astronomy Facility, as an Interdisciplinary Scientist with a proposal to study the Perseus Cluster and other cooling flows. This meant that up and until the launch in 1999, after which AXAF was renamed Chandra, I regularly travelled to the US, particularly to Marshall Space Flight Center in Huntsville, Alabama. I was also shifting the main focus of my research onto clusters of galaxies and Active Galactic Nuclei (AGN). I also carried out committee work in the UK (ASR Board of the Space Engineering Research Center) and European Space Agency (AWG and future planning with Horizon then Horizon Plus). In 1979, I joined the editorial board of Monthly Notices of the Royal Astronomical Society dealing with papers in X-ray and Gamma-ray astronomy up until 2008, being managing editor for the final 14 years). I stopped editing when I became President of the Royal Astronomical Society in 2008.
Why are you doing science?
Research funding in the UK for astronomy had been hit hard and we mobilised an Astronomy Forum which included a senior representative from all UK departments researching astronomy and approached government. I found it easier to talk with Science Ministers than with their civil service advisors and gave a Presidential Address on the Impact of Astronomy, at a time when Impact Factors were the centre of funding discussions. I outlined
“the impact astronomy has had on our society historically, and at present, in terms of cultural, technological and economic benefits. Also why these benefits are so difficult to quantify in terms of the contribution made by basic science. I hoped to show that we all need to do what we can to promote the worth of our work in the wider world, at this difficult time for public spending” (A&G June 2010).
Sometimes it is good to step back and ask why you are doing your science and why someone should pay you to do it! Propagating the scientific method may be a good place to start. My observational work in the early 80s expanded to include optical studies, particularly of clusters cores and the H alpha nebulosities seen there. Theoretically, I explored the possibility of pair plasmas being created around luminous accreting black holes. I picked this up in 2015 in work with Ann Lohfink using data from Fiona Harrison’s NuSTAR observatory.
A new path of relativistic reflection
Stimulated by work by my ex-student and postdoc, Paul Guilbert, with Martin Rees, in which they argued that it was plausible for relatively cold gas to occur in accretion flows close to a black hole, in 1988 I considered the X-ray irradiation of a cold accretion disc in that situation. I realised that a fluorescent iron emission line was likely produced, and that it would be relativistically broadened by the strong gravity effects - doppler and gravitational redshifts - close to the black hole. This could explain a puzzling broad iron line seen by Nick White and others from Cyg X-1 using EXOSAT. I discussed it with Nick and he suggested I contact his colleague Luigi Stella who computed the expected line profiles. Our joint paper written together with Martin was published in 1989 and launched what was for me a new path in relativistic reflection.
Juggling different projects
One of my postdocs, Ian George, and I used Monte-Carlo methods to generate X-ray spectra of X-rays reflected from cold gas. Long-standing collaborator and visitor Randy Ross computed the spectra from gas ionized by the irradiating flux. A theoretical picture had emerged but clear observational evidence was lacking. Ginga spectra from Ken Pounds, Paul Nandra and others showed the expected hard X-ray emission hump from reflection but relativistic effects needed higher spectral resolution. That came after the 1993 launch of the Japanese-US observatory Advanced Satellite for Cosmology and Astrophysics (ASCA), which carried the first charge-coupled device detectors for cosmic X-ray astronomy. I had joined the ASCA team as a science advisor at the generous invitation of Yasuo Tanaka, the Principal Investigator of the mission. My wife, Carolin, and I spent a happy three and a half months in Japan in the summer of 1993 working on ASCA data at Institute of Space and Astronautical Science with other visitors including Richard Mushotzky from Goddard Space Flight Center and MIT student Keith Gendreau. It was an exciting time with many observations yielding new discoveries.
Observations of the bright active galactic nuclei, MCG-6-30-15, showed rapid variability and a strong iron line with a hint of broadening. A long observation was required to substantiate this and was arranged by Yasuo for four days in 1994. A broad iron line with a shape similar to our predictions in 1989 emerged from the spectrum and was published in Nature in 1995. Later that year, I led a paper discussing why alternative origins for the skewed broad shape were either incorrect or implausible. Although instantly accepted by some, it took a long time for others to adopt the model.
At the same time I was working on data from clusters of galaxies, from ASCA and from ROSAT (ROentgen SATellite), which had been launched in 1990. Led by Hans Bohringer, analysis of ROSAT High Resolution Imager data from the centre of the Perseus cluster showed that the double radio source had displaced the hot X-ray emitting gas. The active galactic nuclei was disturbing the inner gas but not necessarily anything else. There was much other work on clusters going on in my group, including the measurement of gas fractions and its implications for the mass fraction of the universe, identification and study of new massive cooling flows clusters from the ROSAT All Sky Survey, and a wide variety of active galactic nuclei phenomena.
Both Chandra and XMM-Newton were launched in 1999 and a fantastic flood of exciting data began. My family and I witnessed the night Shuttle launch of Chandra, when night turned into day. A month later we saw the inverse in the total eclipse of the Sun from Alderney in the Channel Islands. The launch and the eclipse are both ‘whole-body’ events that have to be experienced rather than viewed in photos. We have seen two further total eclipses but no more launches. I spent much of my Chandraguaranteed time looking at Perseus and similar clusters, with Jeremy Sanders and others, which rewarded us with immense detail and improving with subsequent longer exposures until we had about a megasecond of data with 100 million photons by 2006. Eugene Churazov had a model for expanding bubbles generated by the central active galactic nuclei matching the ROSAT image and now seen in considerable detail with Chandra. How energy would be spread widely into the hot gas was unclear until we discovered ripples in 2003. They looked to me like sound waves generated by the bubbles. Whether this is the correct interpretation or not still awaits an even longer exposure.
A vital step
The Reflection Grating Spectrometer (RGS) on XMM clearly showed that although the hot gas dropped in temperature towards the centre of many clusters, it was not radiatively cooling much below ten million K. The energy lost in radiating the X-rays we see was being balanced by heat being supplied, presumably by the central active galactic nuclei. Such clusters became known as cool core clusters and account for about one half of all clusters. The details of the processes involved are still strongly debated. The central black hole controls the behaviour of gas out to a radius a billion times or more its own (event horizon) radius. The overall process is one aspect of black hole feedback and involves jets from matter very close to the black hole blowing bubbles in the surrounding gas.
X-ray spectroscopy capable of measuring the flows of gas in a cool core is a vital step in making further progress. This was taken with the Japan-NASA-ESA observatory satellite ASTRO-H, renamed Hitomi following its February 2015 launch. I was again a scientific advisor eager to look at the spectra expected from the microcalorimeter array, which operated at 50 mK, giving an unprecedented spectral resolution in space of 5 eV. The centre of the Perseus Cluster was the first target and was observed for about 200 ks producing a fantastic spectrum of the emission-line rich intracluster gas. I spent 2 weeks in March at ISAS in Tokyo at the invitation of PI Tad Takahashi working with the Rich Kelley and the Hitomi team on the spectra, revealing that the gas had a mean random velocity of about 160 km/s, with an uncertainty of less than 10 km/s. An amazing result, with a June publication in Nature. This for me emphasised that something more than turbulence was required to transport the active galactic nuclei energy across the core, with sound waves a strong contender. The excitement engendered by the enormous success of this first spectrum was tempered by the loss of the spacecraft, and instruments, a few weeks after the Perseus spectrum was obtained.
I was honoured to give a talk on the results in May at the American Astronomical Society meeting in Naples, Florida. One evening there I saw Hitomi flashing in the dusk sky, reflecting sunlight as it spun rapidly in its orbit and reflected that the path of observational research is not necessarily a straight one.
It has been known for decades that massive black holes are likely to occur in many galaxies: Donald Lynden-Bell discusses dead quasars in his prescient 1969 paper. If gas falls into them they can become very luminous and known as AGN or quasars. In a sense they were seen as more of an ornament in the galaxy centre, almost separate from the rest of the galaxy. In the late 1990s, the black hole mass of a galaxy was found to correlate with the mass of the galaxy (or the bulge part of the galaxy). This sparked the idea, in a 1998 paper by Joe Silk and Martin Rees, that the black hole might control the galaxy, not the other way round. They showed that energy from the black hole can expel gas from the galaxy, stopping star formation and making old galaxies red and dead. A year later I published a paper arguing that momentum was more important for lifting gas out of a galaxy, by analogy with the Eddington limit and the rocket equation which is centred on momentum. A little later I pointed out that this approach led to a relation that agreed well with the observed black hole mass - galaxy mass relation. Radiation pressure on dust might be the active process in such feedback.
I have since continued playing and working on this process with students and postdocs. Wako Ishibashi and I have shown that it can produce outflows resembling those observed and, with Robert Maiolino, that observable stars might form in the outflows. Active galactic nuclei feedback can both stop existing star formation and start new star formation on low angular momentum orbits. It can change the shape of a galaxy.
The action of radiation pressure on dust appears to agree with the column density distribution of absorbing gas in active galactic nuclei as a function of their Eddington fraction. It will be exciting to see whether the shaping of galaxies by active galactic nuclei radiation acting on dust is supported by the ongoing eROSITA X-ray surveys.
Relativistic reflection in active galactic nuclei and black hole X-ray binaries has become commonplace with XMM observations and more recently with NuSTAR and NICER. Relativistic light bending has been needed to explain our XMM data as shown in work with Giovanni Miniutti. Jon Miller showed that reflection is common in X-ray binaries. Reverberation, which was mentioned at the end of the 1989 paper was first spotted in our XMM data of 1H 0707-491 by Phil Uttley. It has been explored further by my students Abdu Zoghbi and Erin Kara and by postdoc Will Alston, as well as by others. It is a strong confirmation of the relativistic reflection interpretation. When 1H 0707 dropped into a low state I found that the spectrum was best interpreted as originating from within two or three gravitational radii around a rapidly spinning black hole. Early work with Kazushi Iwasawa and Anthony Lasenby in the 1990s suggested that we see evidence for black hole spin from the small disc inner radii inferred from spectral fits of broad lines, including MCG-6. Chris Reynolds and Laura Brenneman have more recently systematised this and shown that we have a powerful tool for measuring spin. Together we have explored the observing systematics taking into account the spin dependence of radiative efficiency of accretion and shown that current flux-limited surveys favour rapidly spinning objects.
Recent work with Javier Garcia and others has led to computations and testing of high-density reflection, matching the conditions expected in X-ray binaries and lower mass active galactic nuclei. Much more can be done with current instrumentation such as XMM + NuSTAR but our work is basically photon-starved, particularly in the case of reverberation studies. It is true for both luminous accreting black holes and cool core clusters. It has been great fun and very productive to look at bright Galactic sources with Keith Gendreau’s NICER on the International Space Station.
I was very fortunate to have an ERC Advanced Grant to fund my group from 2013-2018. These generous European awards enable a strong focus on the best science for a five-year period and I was able to build a very strong group working well with each other, my students and myself on the topic of active galactic nuclei Feedback. We covered many of the topics mentioned above and benefitted from successful observations with a variety of satellites and telescopes. We exceeded the “critical mass for which the whole exceeds the sum of its separate parts.” Most of the postdocs from the group have been awarded fellowships or faculty positions.
A tangled web
I have thoroughly enjoyed my life as an observational X-ray astronomer. Most of the work has been in collaboration with others and I have benefitted greatly from their teaching, mentorship, discussion, hard work and humour. The international spread of collaborators is enormous and highly beneficial. I like to work through the simple theoretical aspects and explanations of the objects and phenomena we have observed. That has often led to further ideas and observations. Although my work has covered a very wide range of objects, I see useful interconnections throughout. A talk on stars or planets can stimulate ideas on quasars and galaxies. There is a tangled web both in the physics and in the interactions with students, postdocs and collaborators.
I’m a firm believer in the value of serendipity in science in the Pasteur sense of “chance favours the prepared mind.”
"I tell my students that I am helping to prepare their minds. I also tell them that I do two things for them: one is to start them off with some good ideas that are do-able but have not yet been done, the other is to tell them when they’ve done enough on a project, since all projects are semi-infinite."
I also like Harwit’s concept of Discovery Space in which the coordinates are space, time, resolving power, collecting area, wavelength, etc. When we look tenfold deeper in any parts of this space then we are likely to discover something really new. I have seen this happen again and again. The success of our telescopes is often measured in terms of new things they discover, yet the proposals for building those telescopes depend on what is already known and how much better that can be measured. There is a tension here that is unresolved. It is also a tension in observing proposals for using a telescope in that you rarely win time by arguing that you just want to look deeper without stating clearly what you will find or test. Maybe there is a parallel here to Churchill’s quote on democracy? (No one has come up with a better method of proposal selection). In 2013, I was part of a small team, led by Paul Nandra, that wrote a successful proposal to ESA for the Athena mission, a billion euro orbiting X-ray observatory for studying the hot and energetic Universe. I became a member of the Science Study Team and again made many visits to ESteC, over 40 years after my visits for SL91. Athena is due for launch in the early 2030s, and I look forward to learning of the new discoveries it has made. That is if I last that long!
I am very grateful to my many students, postdocs and collaborators for working and exploring the Universe with me and to Roderick Johnstone and Judith Moss for long-term support. I am indebted to Carolin for love, support and companionship. We share a deep interest in astronomy. Our (biologist) sons, Sam and Laurie, have tolerated many overseas trips with us to the extent that at one stage they preferred to drive to Snowdonia rather than fly to California. Now they view us with bemused good humour. The coronavirus lockdown means that Carolin and I continue to explore the night sky with a small telescope in our back garden. There are always new things to see and new ways to see them.