Figure 1: A Subaru Suprime-Cam image for one of the clusters used in the analysis, A2390 (2.7 billion light years from Earth). The purple hue shows the dark matter distribution measured by the gravitational lensing effect on distant galaxies (typically 8 billion light years from Earth), with the darker color indicating the denser dark matter concentration. It shows that the dark matter distribution is elongated along the northwest-southeast direction.
Figure 2: An illustration of the measurement of the dark matter distribution using gravitational lensing. Color contours indicate the density of dark matter, with redder color marking higher density. Black ellipses show the distortion pattern of background galaxies; positions of distant galaxies are systematically distorted into the shapes shown by black ellipses due to the gravitational lensing effect. (In practice, background galaxies have their own shapes and orientations, and hence we average many galaxies' shapes to extract distortions by gravitational lensing). Left and right panels show spherical and elliptical distributions of dark matter respectively. The difference in the distortion patterns suggests that one can measure the shape of the dark matter distribution from two-dimensional lensing distortion patterns.
Figure 3: The number distribution of the "ellipticity" of the dark matter distribution in 18 clusters of galaxies, obtained from the team's analysis of the gravitational lensing effect seen in Subaru images. Zero ellipticity means that the distribution is spherical, while a larger value of ellipticity indicates a more flattened distribution. The measurement peaks at around 0.5 (corresponding the 2:1 ratio of major to minor axes of the ellipse), suggesting that the flattened shape of the dark matter distribution is detected at a high level of significance. The black line shows a theoretical prediction based on the standard collisionless, cold dark matter model, made in 2002 by Yipeng Jing (Shanghai Astronomical Observatory) and Yasushi Suto (University of Tokyo). This demonstrates that the observed distribution agrees well with the theoretical prediction.
The nature of dark matter is still unknown and is currently a central problem in modern astronomy and physics. Dark matter is dark in a couple of ways. It is undetectable to visible light and has escaped detection at all electromagnetic wavelengths. Because it is invisible, its existence has to be inferred from its gravitational effect on other celestial objects as well as from theoretical models. Indirect evidence has established its relative abundance in our universe-probably five times greater than visible matter-in addition to its significance for understanding galaxy formation. For example, a considerable amount of dark matter probably sustains the structure of galaxies, because the gravitational force of visible matter cannot bind its member stars. The scientific challenge is how to study the nature of dark matter. Astronomers seek ways to use their observations to solve this puzzle.
One approach to a solution is to make detailed measurements of the spatial distribution of dark matter and then compare the data to predictions drawn from theoretical models. Both aspects of this approach have their difficulties. How can the distribution of dark matter be measured? What are plausible assumptions to include in models of dark matter?
A team of astronomers led by Masamune Oguri at the National Astronomical Observatory of Japan and Masahiro Takada at University of Tokyo decided to use gravitational lensing to measure and analyze the distribution of dark matter. Gravitational lensing provides a unique opportunity to explore dark matter distributions by measuring the distances that light travels from distant to foreground objects. Einstein's general theory of relativity predicts that light from a distant object will bend around a massive object in the foreground, e.g., a cluster of galaxies or a concentration of dark matter. By measuring the distortion pattern of many distant galaxies, it is possible to infer the mass(es) of the object(s) in the foreground. Since the technique does not rely on assumptions about the visibility of the matter bending the light, gravitational lensing can be a powerful probe of dark matter.
The team fine-tuned their research by observing 20 massive clusters of galaxies with the Subaru Telescope's Prime Focus Camera (Suprime-Cam). Clusters of galaxies are ideal sites for studying the distribution of dark matter, because they contain thousands of galaxies and are known to accompany a large amount of dark matter. The superb light-collecting power and excellent image quality of the Subaru Telescope gave the researchers an extra advantage. By using Suprime-Cam at prime focus, they could capture objects in a particularly wide field-of-view.
Observations with Suprime-Cam yielded wide-field images of 20 massive clusters of galaxies (typically located at 3 billion light years from Earth), which the team then used to measure and analyze dark matter distributions (Figure 1). From their detailed analysis of gravitational lensing effects in the images, the team obtained clear evidence that the distribution of dark matter in the clusters has, on average, an extremely flattened shape rather than a simple spherical contour (Figure 2 and 3). The measured degree of the flattening is quite large, corresponding to 2:1 in terms of the ratio of major to minor axes of the ellipse. This finding represents the first direct and clear detection of flattening in the dark matter distribution with the use of gravitational lensing.
In addition to the promise of using gravitational lensing for exploring the nature of dark matter, this research contributes to the theoretical modeling of dark matter. Detailed comparisons of the team's findings with theoretical model predictions of the distribution of dark matter show that the observed degree of the flattening is in excellent agreement with theoretical expectations.
Theoretical predictions for dark matter distributions in clusters of galaxies are dependent on what kind of dark matter model is assumed. This research strongly supports the prevailing model, which begins with the assumption that dark matter consists of weakly interacting massive particles that are relics of the Big Bang. These particles are assumed to be "cold", i.e., thermal motions of the particles are negligibly small. According to this scenario, clusters of galaxies are dynamically young objects that form through the merging of many small objects. This theory predicts that the dark matter distribution in clusters of galaxies would be non-spherical, reflecting a large-scale structure of dark matter filaments (i.e, ribbons of cold material, see also this release). Since the team's findings confirm a non-spherical distribution, they demonstrate the feasibility of exploring the nature of dark matter via flattening in the dark matter distribution.
This study is a part of the Local Cluster Substructure Survey (LoCuSS), an international project carrying out a systematic study of galaxy clusters by combining the Subaru data with a wide range of data sets from radio, infrared, optical and X-ray telescopes. The goal of the project is to reveal new aspects of cluster physics and cosmology. The results presented here are one of its initial achievements. More detailed explorations of dark matter in clusters of galaxies are planned using a wide-field survey with the Hyper Suprime-Cam, the next generation prime-focus camera on the Subaru Telescope. These future studies will greatly improve our understanding of galaxy clusters and, hopefully, the properties of dark matter.
The results of this research will be published in Monthly Notices of the Royal Astronomical Society.
The List of Authors
Masamune Oguri (National Astronomical Observatory of Japan)
Masahiro Takada (IPMU, University of Tokyo, Japan)
Nobuhiro Okabe (Institute of Astronomy & Astrophysics, Academia Sinica, Taiwan)
Graham P. Smith (University of Birmingham, UK)
