A New Test for Gravity

Those of us who once shared a bed with an older sibling likely gained, in those formative years, an intuitive grasp on the general theory of relativity. Given an adequate difference in the size of the children, the bigger kid sinks deeply into the mattress, creating for the smaller kid an inescapable downward slope toward the often forbidden other half of the bed.

This was, in a sense, Einstein’s picture of the universe: massive objects sink down into the fabric of space-time, creating the gravitational force that pulls lighter objects toward them. In all of the experiments designed to test it, this model of the universe stood up extremely well — at least, until about ten years ago, when astronomers discovered that the expansion of the universe is accelerating rather than slowing down, as had been predicted by Einstein’s general theory of relativity.

Scientists have come up with two possibilities for what’s going on: either there is a huge amount of mysterious, gravitationally repulsive dark energy pushing everything in the universe apart, or gravity simply doesn’t work the same across large portions of the sky as it does on small scales.

That second possibility — that general relativity doesn’t apply on large scales — is the finding of a recent paper by Rachel Bean, a cosmologist at Cornell University.

The paper, which was published in September on the physics website Arxiv.org, has drawn intense interest from the scientific community because it represents one of the first efforts to test general relativity over large spans of the universe. “It’s very provocative,” said Scott Dodelson, a cosmologist at the University of Chicago and the Fermi National Accelerator Laboratory, who was not involved in the study. “I think her results are being taken very seriously.”

At the heart of Bean’s study is data from COSMOS, a deep-field survey conducted with images from the Hubble Space Telescope. Astronomers used weak gravitational lensing — a technique that involves measuring the distortion of light from faraway objects — to gain pictures of large-scale structures in the universe at different points in its history. These pictures tell a story about the evolution and composition of whole clusters of galaxies.

In her research, Bean examined weak-lensing data from three different time slices of large scale structure, and found that in the oldest time slice, general relativity did not accurately predict the way that the structure was behaving. In other words, it didn’t appear to be following the rules of gravity.

Bean was quick to say that these are very preliminary results, which hinge strongly on the validity of the time slice data she used.

Alexie Leauthaud, an astrophysicist at the Lawrence Berkeley National Laboratory who has been working on the weak-lensing data from the COSMOS survey, also emphasized caution about the accuracy of the data. She said improved analysis methods have brought some major changes to the COSMOS data since it was published in 2007 — changes, she thinks, that may influence Bean’s findings.

A similar study to Bean’s, conducted last year by cosmologist Robert Caldwell of Dartmouth College, used weak-lensing data from a different source and found no inconsistencies with general relativity. Caldwell’s study looked at more recent pictures of the universe; it did not extend as far back as the time slice that Bean found to deviate from general relativity.

It remains to be seen whether Bean’s finding will hold with corrected data from the COSMOS team and subsequent re-analysis. Either way, cosmologists agree that her paper demonstrates an emerging ability to probe the mechanisms of the universe at large — an ability that is expected to increase in the coming decade with the implementation of several dedicated telescopes and improved weak-lensing techniques.

If her findings do hold, it wouldn’t mean Einstein’s general theory of relativity has to be thrown out altogether — just that we need a new set of laws to explain how gravity works over large scales in the universe. “It’s not that unusual to have a law that is valid in one domain but supplanted by another description in another domain,” said Robert Caldwell. For example, Newtonian physics describes the behavior of objects at low velocities. But accelerate those objects to a velocity close to the speed of light, and you need Einstein’s special theory of relativity to explain what’s happening.

If Bean’s paper is right, and general relativity doesn’t explain the universe on large scales, “we may just be discovering new laws of gravitation,” said Caldwell.