Ordinary Meeting, 2004 March 31
Illuminating the Invisible: Uncovering Dark Matter with Gravitational Lensing
The speaker commenced with a Hubble Space Telescope (HST) image of galaxy cluster Abell 2218, a useful starting point as it illustrated two quintessential observations upon which her field was built. The first was that mass bends light; the second was the existence of dark matter. Within the cluster could be seen a number of galaxies, all of similar colour because they had formed together and were at similar stages along their evolutionary path. The speaker's first observation, that mass bends light, was derived from a number of arcs of much bluer colour, which appeared to be lensed images of a more distant galaxy. The latter deduction, that dark matter was required to explain the observed phenomenon, derived from the inability, by a substantial factor, of the estimated amount of visible luminous mass in the galaxy cluster to account for the magnitude of the observed lensing of the background galaxy. Dr Gray drew analogy between the HST view and an image of the Earth by night. Just as the latter revealed only those street-lit urban areas, the rest being shrouded in darkness, so our telescopes, sensitive only to glowing luminous matter, gave an incomplete image of the Universe.
The speaker explained that the term 'dark matter' on occasion caused confusion, for it was only dark in its non-emission of electromagnetic radiation. It might be more appropriately named 'transparent matter' in contrast with the dust lanes of dark nebulae, which blocked the passage of light, thus appearing dark. By contrast, dark matter was simply invisible. But, however ambiguous, the name 'dark matter' was here to stay.
Dr Gray explained that her talk would firstly summarise what we knew about dark matter, and how we had arrived at the conclusions we had. She would then provide an introduction to General Relativity as applied to gravitational lensing, before concluding with unanswered questions. The first evidence for dark matter, she explained, had arisen in 1933 when Fritz Zwicky, at Caltech, observed that galaxy clusters orbited around their centres more rapidly than could be explained. Unless there was significant mass in the cluster over and above what was visible, the cluster would fly apart. Extra mass was needed to act as gravitational "glue". Further observational evidence for dark matter would emerge in the 1970s, with the measurement of the rotation velocities of disc galaxies. In most cases, these too were also found to rotate faster than could be explained, particularly towards the edge of the disc. To explain the observed rotation profile, the galaxy had to sit in a halo of invisible, but gravitating, mass.
From such studies, astronomers had been able to pin down how much dark matter there was in the Universe: roughly 8kg for every kilogram of luminous matter. Thus, astronomical observations told us that, in total, dark matter significantly outweighed all the luminous matter, but was much more evenly spread out in space. However, the question of what dark matter was composed of was not one astronomers could answer; this was in the realm of particle physics. Formerly it had been thought that dark matter might be ordinary (baryonic) matter, possibly in the form of small planets, dim stars, black holes, or massive compact halo objects (MACHOs) in the outskirts of galaxies. However, in recent times, cosmologists, studying the initial proportions of hydrogen and helium which emerged from the furnace of the Big Bang through a process called nucleosynthesis, had been able to place stringent constraints on the amount of baryonic matter in the Universe which were compatible with their observations. The upper limit was less than 10% of the amount of dark matter required to fit the rotation velocity observations, and this led them to conclude that most dark matter must be of an exotic kind.
The speaker explained that this led to the presumption that dark matter was composed of a sea of invisible lone particles called weakly interacting massive particles (WIMPs), whose influence could only be felt by their gravitation. Particle physicists already anticipated such particles might exist from a theory known as supersymmetry, and had built a detector in a potash mine in Boulby, Yorkshire, where finely controlled conditions underground might allow the presence of WIMPs to be detected in the next few years.
Dr Gray moved on to discuss the history of our theoretical understanding of gravitational lensing, a story which could be traced back to Newton. His Theory of Gravity suggested that light might be attracted towards massive bodies in the same way as matter. In a moment of foresight, he proposed that light rays passing a mass such as the Sun might be bent towards it, though the effect was so small that it was to remain untested until 1919, four years after Einstein had proposed a new theory of gravity, General Relativity. This made the concrete prediction that the effect should be twice as large as predicted by Newton. Sir Authur Eddington realised that the total solar eclipse of that year would provide a unique opportunity to test the new theory. By noting the apparent positions of stars at small angular separations from the Sun, normally invisible in its glare, he hoped to measure the bending. In a hastily prepared experiment, he compared the positions of these stars on the day of the eclipse as compared to six months earlier, when the Sun had been elsewhere in the sky. Despite poor observing conditions, Eddington declared an observed shift of 1"75, matching exactly Einstein's prediction. Subsequent re-examination of Eddington's plates, however, had shown the data to be questionable, and it was somewhat serendipitous that his conclusion would later be verified by more accurate studies.
The speaker described how Fritz Zwicky, already mentioned for having discovered the first evidence for dark matter, would go on to predict with great accuracy the applications of the lensing induced by it. In 1936, he foresaw the use of gravitational lenses as telescopes to view magnified images of more distant objects, and also predicted that lensed images could be analysed to map the distribution of dark matter. These turned out to be accurate prophecies as to the focus of research sixty years on. In addition, the phenomenon had also been utilised to constrain cosmological parameters. Zwicky's legacy was especially impressive since the first observation of dark matter lensing would not come until 1979: he was truly a man ahead of his time. The discovery would come when Walsh et al. stumbled serendipitously upon what appeared to be a pair of quasars at very close separation. However, these being rare objects, it was statistically unlikely that two should lie so close together, and moreover, when their spectra were found to be indistinguishable, the mystery deepened. These were, they eventually concluded, two images of the same object, traversing different paths around an intervening lens. Suddenly lensing was no longer confined to pure theory.
The behaviour of gravitational lenses could be divided into several classes, Dr Gray explained, and each had its unique applications. The smallest class produced an effect termed microlensing, where a compact stellar mass object within our Galaxy magnified the observed image of a more distant star as it passed in front of it. To terrestrial observers, the star would appear to brighten with a characteristic lightcurve, quite distinct from any stellar variability. While this possibility had been recognised for decades, the exactness of the required alignment meant that the occurrence would be rare and brief, perhaps one event per star in a million years, and prospects for observation were slim. With the recent advent of automated searches, however, it was now possible to observe many millions of stars on a regular basis, introducing the possibility of picking up a handful of events per year. Over the previous decade, the MACHO collaboration had undertaken such a project with many successful detections. The most interesting results from this had been information obtained about the lensing objects themselves, which often turned out to be very peculiar bodies. One recent event appeared to have been caused by a double star with a planet, which would theoretically appear so unstable as to have negligible lifetime.
The speaker next discussed lensing by entire galaxies, weighing around 1011 solar masses, of more distant galaxies or quasars. Termed strong lensing, this could precipitate three effects upon the image of its subject: magnification, distortion, or multiple images. One of the finest examples of all three was the so-called Einstein Cross (2237+030), in which four images of a quasar were seen around the low-redshift lens galaxy. In cases of near-exact alignment, a continuous halo-like ring around the lens galaxy could result, termed an Einstein Ring. In systems with multiple images, the curious possibility arose that the light path-length for the images might vary considerably, and if the lensed source exhibited any sudden variability in its luminosity, this difference could be measured. Indeed, such a phenomenon had now been successfully observed, and, in at least one case, a putative time offset of several years was identified between two of the images, implying a path difference of several lightyears.
The exciting potential of strong lensing to behave as a "telescope", providing a magnified view of very distant objects, had begun to unfold with the discovery on 1997 July 30 of an image of a galaxy apparently at redshift 4.92. In the weeks preceding the meeting this record for the most distant object had been swept away by the claimed discovery of a lensed image of a redshift 7 galaxy by Kneib et al. on 2004 February 12. But this was only to be superseded by a claim of a redshift 10 discovery on 2004 March 1 by Pello et al. using the Very Large Telescope (VLT) in Chile.
Finally, Dr Gray discussed the subtler effects of lensing by more diffuse large-scale mass distributions such as clusters of galaxies, termed weak lensing. If a background field of galaxies were viewed through such a lens it would appear warped or sheared, and the distortion would be constant over larger angular scales than for strong lensing by smaller objects. It might lead to a stretching of the image of every background galaxy in a certain direction, for example. In that case, without any knowledge of the physical orientation or ellipticity of the galaxies, it was impossible to detect that any one galaxy appeared stretched. However, taking a field containing a large sample of galaxies, one would expect each to be randomly aligned. By averaging over their alignments, one could be pretty sure that weak lensing was at work if there was a significant bias towards a particular orientation. Typically, work in this field involved simulating the lensing of a hypothetical field of randomly aligned galaxies, and varying the lensing mass distribution until the statistical properties of the resultant image matched the observation. Thus, despite the subtlety of the effect, the underlying dark matter distribution could be discerned with some accuracy.
To close, the speaker outlined the unanswered questions of the field, including most fundamentally what dark matter was composed of. Also, it was not known how dark matter was distributed through the Universe, or how this had evolved throughout its cosmic history. Finally, the effect of dark matter upon the formation and evolution of galaxies was poorly understood. Following the applause, the President invited questions. When asked, with reference to rotation curves, why she had said "most" galaxies appeared to contain dark matter, rather than "all", the speaker explained this was a difficult problem. There did appear to be a small number of galaxies with no evidence for dark matter, but this could not theoretically be explained, and needed further investigation. In a vote of thanks, the President congratulated the speaker for clearly explaining difficult maths, providing a very illuminating talk. Following further applause, the speaker invited Mr Martin Mobberley to deliver his regular Sky Notes.