IT STANDS out like a neon sign in the centre of the image, a thin blue arc floating over a golden-red blob. The blob is a giant elliptical galaxy six billion light years from Earth in the constellation Fornax. The arc above it is an illusion – a cosmic mirage created as the light from a young blue galaxy up to 10 billion light years away is dragged into a semicircular shape by the golden-red giant’s gravity.
Captured in 2003 by the Hubble Space Telescope, this is just one of many stunning gravitational lens images that decorate the sky. Gravitational lenses are nature’s giant telescopes, giving glimpses of what lies beyond the edge of the visible universe.
But look at them in the right way and these shining, candy-coloured arcs offer more than just distorted snapshots of young galaxies. They can also tell us about the age of the cosmos and the strange dark matter that is sprinkled throughout it.
Advertisement
In the past few years, some astronomers studying cosmic lenses have begun to do just that. Their results have led them to throw down the gauntlet to their colleagues, pitting one cherished idea against another. One of these ideas is the apparent age of the universe. The other is our understanding of dark matter. The results say one of them must be wrong.
Leading the challenge to received wisdom is Christopher Kochanek, who studies cosmic lenses at Ohio State University in Columbus. Together with Paul Schechter of MIT and others, Kochanek analyses lenses that produce two, three, or more images of distant galaxies and quasars. The science he uses dates back to 1911, when Albert Einstein showed how the presence of a mass bends light rays that pass it.
According to Einstein’s general theory of relativity, a large mass like that of a giant galaxy will distort space-time. The idea of distorted space-time is often compared to the way a bowling ball might make the elastic fabric of a trampoline sag. If a smaller ball rolls across the trampoline, its path will bend around the larger object. In the same way, light rays bend when they travel through the regions of curved space around massive objects like galaxies.
Gravity thus acts as a lens, deflecting the light from distant stars or galaxies. To an observer on Earth, light from these distant objects appears to come from different parts of the sky than it really does, producing magnified, ghostly images of celestial objects.
“It would force an upwards revision in the age of the universe from 14 billion to 20 billion years”
If a distant quasar, a large galaxy and the Earth are exactly aligned, the bending of light by the galaxy will cause the quasar to appear as a smeared-out ring – an Einstein ring – around the lensing galaxy (see Graphic). But if the three are not aligned, and the lensing galaxy is not symmetrical, the cosmic mirage can look hopelessly complicated. It takes careful study by astronomers like Kochanek to work out precisely how several strangely shaped objects in the sky can be multiple images of the same distant quasar. It becomes even more complicated because quasars change over time, alternately brightening and dimming. When a quasar flashes, so do its lensed images, but they change at slightly different times, because the light rays producing the different images take different paths through space to reach Earth.
It’s this time lag that can be used to pin down the age of the cosmos. By showing how much longer one path through space is compared with another, the time delay reveals how far apart the lens galaxy and the distant quasar are. And these distances can be used to pin down the most important number in cosmology, the Hubble constant.
We live in an expanding universe in which galaxies are receding from each other. Think of the universe as a rising loaf of raisin bread. Each galaxy is a raisin that is moving away from the others as the loaf expands. The Hubble constant measures how fast the universe is expanding (see “Hubble trouble”). It also tells us the age of the universe. The faster the universe is expanding, the less time it must have taken for it to reach its current size.
Many cosmological observations agree on the constant’s value: around 70 kilometres per second per megaparsec. If our galaxy is one raisin in the bread, then another raisin 1 megaparsec from us is racing away at a speed of 70 kilometres per second. Trace this back and the universe must be 14 billion years old.
But Kochanek and others have measured 10 cosmic lenses precisely enough to challenge this measurement. Four out of the 10 are in striking disagreement with the standard value for the Hubble constant. Instead of 70, these lenses give a value of around 50. This would force an upward revision in the age of the universe, from 14 billion to 20 billion years. It’s a huge difference, and if the new figure is true it would demand an overhaul of everything we think we know about the cosmos.
But is that justified? The time delay studies depend critically on knowing exactly how the lensing galaxy distorts the path of light. The distribution of mass in the lens galaxy affects the path of light rays from the quasar to Earth, so the model has to be incredibly detailed. “To model well, we need a good knowledge of the source and the lens,” says Charles Keeton, who works on lensing at Rutgers University in New Brunswick, New Jersey. “Our model should predict the positions and brightness of the seen images. We make a model, determine the lens properties it would have, and then adjust its parameters to fit what is actually seen.”
“We need to know the relative positions of the images and the lens, the source position, the time delay between images and the relative brightness of the images,” says Edwin Turner at Princeton University in New Jersey. In other words, a lot can go wrong in arriving at the sort of result Kochanek has obtained.
The most important assumption has to do with the distribution of mass in a lens galaxy, because it is the mass that does the lensing. While the position of the stars and hot gas in a lens galaxy can be seen clearly by telescopes such as Hubble, this matter accounts for only a fraction of the mass of the galaxy. The rest is dark matter, an exotic form whose presence is inferred from its gravitational influence on bright, visible matter, but which stubbornly defies direct detection.
Astronomers know roughly how much dark matter is out there because its gravity holds together rotating galaxies that would otherwise fly apart. After making hundreds of observations, they have a standard theory that predicts exactly how this dark matter should be spread in galaxies. Based on the theory, the bright matter of a galaxy is believed to be embedded in a huge halo of dark matter. Researchers who study dark matter have no reason to think dark matter would be clustered any other way.
Stripped of dark matter
And this is exactly what Kochanek’s calculations assume. “In its simplest form, the lens galaxy is modelled as an ellipse,” says Keeton. Good lens models take into account the shape of the lens, the spread of dark matter in it and the influence of neighbouring bodies on it. And these three contributions are added on to the simple model.
But supporters of the higher, standard value for the Hubble constant believe this is where things can go awry. “The disagreement relates to how the lens studies are done,” says Josh Winn of the Harvard-Smithsonian Center for Astrophysics, who has worked on lens modelling with Kochanek.
Suppose for a moment that the Hubble constant is indeed around 70, not the 50 that the lenses seem to be saying. Cosmic lenses could instead be telling us something new about dark matter in galaxies. If Kochanek’s models strip galaxies of the dark matter halos, he gets a completely different figure for the Hubble constant, around 71, slap bang on the standard figure.
Yet some researchers, including Kris Stanek of the Harvard-Smithsonian Center for Astrophysics, question the Hubble constant, and with it the known age of the universe. There are so many studies supporting the idea that galaxies have a lot of dark matter around them, says Joel Primack, who models dark matter halos at the University of California, Santa Cruz. “Most theoretical cosmologists would confirm that the dark matter model fits well with all available data and that new data coming in only serves to confirm its robustness. We are confident we have the right story,” he says.
But most astronomers would argue it is dark matter, not the age of the universe, that should give. Quiz a hundred astronomers about the value of the Hubble constant, says Schechter, and most will pick a standard value, not the one given by lensing studies.
Kochanek says it is difficult to reconcile both entrenched viewpoints. “Modelling galaxies with a mass distribution that includes a dark matter halo leads to an anomalously low value of the Hubble constant from gravitational lenses,” he says. “Raising the value rids us of the dark matter halo. We thus face a conundrum – do we accept the low value of the Hubble constant? Or is there less dark matter than you expect in the lens galaxies?”
One way out might be to say that the lens galaxies are special in some way. The four raising the problem are elliptical galaxies. “It could be that elliptical galaxies may exhibit a variety of mass distributions,” suggests Winn. “Some may follow the model, some, like the lens galaxies, may not.” Keeton agrees: “We could be looking at a sub-section of elliptical galaxies with varying mass profiles.” He points to studies on planetary nebulae, clouds formed when a star expels its outer layers. Some suggest galaxy halos are not always as massive or as concentrated as expected.
Yet there should be nothing special about these lens galaxies. They are simply galaxies that happen to be passing between Earth and a source. “It is reasonable to say that we either don’t understand how dark matter is distributed or we don’t know the Hubble constant well,” says Stanek. “The consistently low results seem to be telling us something.”
One way to solve the impasse might be to take a fresh look at how dark and bright matter interact. “We still do not understand well how luminous matter interacts with the dark component,” says Primack. For example, interactions with bright matter may change the distribution of the dark matter in some unexpected way.
So far, Kochanek hasn’t managed to convince others that cherished ideas need overhauling. Many astronomers prefer to question the lensing results instead. Complications could arise if other galaxies nearby affected the lensing. Kochanek is currently working to create a sample of 20 “clean” lenses that can be modelled accurately, but he believes cosmic lenses are so powerful that what is seen through just one clean lens focused on the edge of the universe cannot easily be ignored.
Hubble trouble
The rate of expansion of the universe is captured by a number called the Hubble constant. To calculate it, astronomers need to know both the distances between us and nearby galaxies, and their speeds as they hurtle away from us. They discover their speed by looking at the colour of light coming from the galaxies – the faster a galaxy is moving away from us, the redder it appears, and the further characteristic lines in its spectrum of light are shifted towards the red end. But getting the distances is much harder.
In the late 1990s, as part of the Hubble Space Telescope Key Project, Wendy Freedman of Carnegie Observatories in Pasadena, California, and others studied hundreds of pulsating stars called Cepheids, which are found in many galaxies. The apparent brightness of any celestial object declines the farther away it is. But cepheids flash at a rate that depends on their actual brightness, so by comparing this with their apparent brightness, they can be used to calculate the distance to the galaxies in which they lie.
When asked about the Hubble constant, most astronomers will quote the Key Project’s value of around 70 kilometres per second per megaparsec, meaning that a galaxy 1 megaparsec (3.26 million light years) from us is moving away at a speed of 70 kilometres per second. “The Key Project number is the widely accepted value,” says Wayne Hu of the University of Chicago.
Yet cosmic lenses can look far further into the past, and some seem to see a smaller Hubble constant than the Key Project found, causing some controversy. “Lensing provides an independent check on the value of the Hubble constant,” claims Kris Stanek of the Harvard-Smithsonian Center for Astrophysics.