Gravity's Role as a Lens and a Cosmological Force

Tyson and his colleagues were energetically conducting their surveys of this blue "curtain" even as other astronomers were discovering faint blue arcs, thought to be the result of gravitational lensing. The blue curtain is a much more efficient probe of foreground dark matter because the curtain is far enough away to enhance the chances of foreground lensing and rich enough to provide a canvas—rather than a point of light coming from a quasar—to observe the distortion. After subtracting the gravitational effects of those luminous galaxies in the foreground that they could observe, Tyson's team could nonetheless still observe fairly dramatic distortion in the blue curtain galaxy images. They realized they had taken an image—albeit an indirect one—of the dark, nonluminous matter in the foreground cluster of galaxies (Figure 4.2). With quasar studies, the foreground object was often a very specific galaxy or star, and the light was emanating from the discrete quasar source. Tyson's system works by the same principles of gravity, but instead of an identifiable body that one can see, the only evidence of the foreground object is the gravitational pull it exerts on the path of the distant light and the resulting tell-tale distortion of the distant galaxy's image (Figure 4.3).

By enhancing FOCAS and developing other programs to simulate

Figure 4.2 The gravitational effect of dark matter in a foreground cluster of galaxies. The images of the foreground galaxies have been eliminated, revealing faint arcs. The clump of dark matter is 4 billion light-years away.

the possible shape and composition of dark matter, Tyson has been able to use his blue curtain as a vast new experimental database. Characteristic distortions in the previously too faint images have become observable as the deep CCD optical imaging surveys and analyses continue. A portrait of dark matter is emerging that allows other theories about the evolution of the universe to be tested—galaxy cluster formation, for one. "We have found that there is a morphological similarity between the distribution of the bright cluster galaxies and the distribution of the dark matter in the cluster, but the dark matter is distributed relatively uniformly on the scale of an individual galaxy in the cluster," said Tyson. Their lens technique is better at determining the mass within a radius, to within about 10 percent accuracy, than the size of the inferred object. "Less accurate," added Tyson, "is the determination of the core size of this distribution. That is a little bit slippery, and for better determination of that we simply have to have more galaxies in the background. But

Figure 4.3 A diagram showing the bending of a light path by the mass in a gravitational lens and the apparent change in the position of the source and the resulting distortion of its image. 

since we are already going to around 30th magnitude or so in some of these fields, that is going to be hard to do."

"A very interesting question that is unanswered currently because of the finite size of our CCD imagers is how this dark matter really falls off with radius. Is the dark matter pattern that of an isothermal distribution of particles bound in their own gravitational potential well, in which case it would go on forever?" Tyson wondered. This is, in part, the impetus for Tyson to develop a larger CCD device.

Results from Tyson's work may aid astronomers in their quest to understand how galaxies formed. In order to create the lumps that eventually developed into galaxies, most theories of galaxy formation start with the premise that the Big Bang somehow sent a series of waves rippling through the newly born sea of particles, both large and small fluctuations in the density of gas. According to one popular theory of galaxy formation, small knots of matter, pushed and squeezed by those ripples, would be the first to coalesce. These lumps, once they evolve into separate galaxies, would then gravitationally gather into clusters and later superclusters as the universe evolves. This process has been tagged, appropriately enough, the "bottom-up" model.

Conversely, a competing theory of galaxy formation, known as the "top-down" model, essentially reverses the scale of collapse, with the largest, supercluster-sized structures collapsing first and most rapidly along the shorter axis, producing what astrophysicists call a pancake structure, which would then fragment and produce cellular and filamentary structures.

Bertschinger explained that gravitational instability causes "different rates of expansion depending on the local gravitational field strength. Gravity retards expansion more in regions of above-average density, so that they become even more dense relative to their surroundings. The opposite occurs in low-density regions."

Following earlier computer models simulating the evolution of a universe filled with cold dark matter, especially a model pioneered by Simon White and his associates in the 1980s, Bertschinger and his colleagues have explored how dark matter may cluster and how this development may parallel or diverge from galaxy clustering: "The gravitational field can be computed by a variety of techniques from Poisson's equation, and then each particle is advanced in time according to Newtonian physics." The complexity comes in when they try to capture the effects of the explosive microsecond of Guth's inflation era. For this they need to model the quantum mechanical fluctuations—essentially the noise—hypothesized to extend its impact through time by the medium of acoustic waves. In their model, said Bertschinger, "most of the dark matter does end up in lumps associated with galaxies. The lumps are something like 1 million light-years" in extent, but it is not conclusive that they correspond to the haloes found beyond the luminous edges of galaxies. "It is plausible," he continued, "that luminous galaxies evolve in the center of these lumps," but to say so more definitively would involve simulating the dynamics of the stellar gases, which was beyond the scope of their model. But dark matter does seem to cluster, and in a manner similar to galaxies. Most of it is found within 1 million light-years of a galaxy. These conclusions are tentative, in that they emerged largely from simulations, but they are consistent with many of the observed data, including the limits found with respect to the anisotropy of the cosmic microwave background.