Astrophysicists led by the University of Chicago's Andrey Kravtsov have resolved an embarrassing contradiction between a favored theory of how galaxies form and what astronomers see in their telescopes.

Astrophysicists base their understanding of how galaxies form on an extension of the big bang theory called the cold dark matter theory. In this latter theory, small galaxies collide and merge, inducing bursts of star formation that create the different types of massive and bright galaxies that astronomers see in the sky today. (Dark matter takes its name from the idea that 85 percent of the total mass of the universe is made of unknown matter that is invisible to telescopes, but whose gravitational effects can be measured on luminous galaxies.)

This theory fits some key data that astrophysicists have collected in recent years. Unfortunately, when astrophysicists ran supercomputer simulations several years ago, they ended up with 10 times more dark matter satellites-clumps of dark matter orbiting a large galaxy-than they expected.

"The problem has been that the simulations don't match the observations of galaxy properties," said David Spergel, professor of astrophysics at Princeton University. "What Andrey's work represents is a very plausible solution to this problem."

Kravtsov and his collaborators found the potential solution in new supercomputer simulations they will describe in a paper that will appear in the July 10 issue of the Astrophysical Journal. "The solution to the problem is likely to be in the way the dwarf galaxies evolve," Kravtsov said, referring to the small galaxies that inhabit the fringes of large galaxies.

In general, astrophysicists believe that formation of very small dwarf galaxies should be suppressed. This is because gas required for continued formation of stars can be heated and expelled by the first generation of exploding supernovae stars. In addition, ultraviolet radiation from galaxies and quasars that began to fill the universe approximately 12 billion years ago heats the intergalactic gas, shutting down the supply of fresh gas to dwarf galaxies.

In the simulations, Kravtsov, along with Oleg Gnedin of the Space Telescope Science Institute and Anatoly Klypin of New Mexico State University, found that some of the dwarf galaxies that are small today have been more massive in the past and could gravitationally collect the gas they need to form stars and become a galaxy.

"The systems that appear rather feeble and anemic today could, in their glory days, form stars for a relatively brief period," Kravtsov said. "After a period of rapid mass growth, they lost the bulk of their mass when they experienced strong tidal forces from their host galaxy and other galaxies surrounding them."

This galactic "cannibalism" persists even today, with many of the "cannibalized" dwarf galaxies becoming satellites orbiting in the gravitational pull of larger galaxies.

"Just like the planets in the solar system surrounding the sun, our Milky Way galaxy and its nearest neighbor, the Andromeda galaxy, are surrounded by about a dozen faint 'dwarf' galaxies," Kravtsov said. "These objects were pulled in by the gravitational attraction of the Milky Way and Andromeda some time ago during their evolution."

The simulations had succeeded where others had failed because Kravtsov's team analyzed simulations that were closely spaced in time at high resolution. This allowed the team to track the evolution of individual objects in the simulations. "This is rather difficult and is not often done in analyses of cosmological simulations. But in this case it was the key to recognize what was going on and get the result," Kravtsov said.

The result puts the cold dark matter scenario on more solid ground. Scientists had attempted to modify the main tenets of the scenario and the properties of dark matter particles to eliminate the glaring discrepancy between theory and observation of dwarf galaxies. "It turns out that the proposed modifications introduced more problems than they solved," Kravtsov said.

The simulations were performed at the National Center for Supercomputer Applications, University of Illinois at Urbana-Champaign, with grants provided by the National Science Foundation and the National Aeronautics and Space Administration.