麻豆传媒

A controversial approach to gravity that challenges Albert Einstein and suggests dark matter doesn鈥檛 exist has passed its first test.

The vast majority of physicists agree that gravity acts according to rules laid down in Isaac Newton鈥檚 law of gravitation and Einstein鈥檚 theory of general relativity. Yet observations of the universe show that the motion of the galaxies can鈥檛 be explained by the gravitational pull of all the ordinary matter out there 鈥� hence the belief in unseen, dark matter that exerts its own pull.

Now, a team of astronomers studying the distribution of matter in more than 30,000 galaxies say their observations can be explained by an alternative theory that does away with dark matter. If this 鈥渕odified gravity鈥� is correct, it would up-end hundreds of years of fundamental physics.

at Leiden University, the Netherlands, and her colleagues looked at the gravitational lensing of these galaxies 鈥� the way they bend the light of more distant galaxies as predicted by Einstein鈥檚 theory 鈥� to measure their dark matter content.

To their surprise, they discovered the observed lensing could just as readily be accounted for by a new model of gravity, without invoking dark matter.

, a theoretical physicist at the University of Amsterdam in the Netherlands, has been developing a competing model of gravity that borrows heavily from quantum mechanics, relativity, information theory and string theory. It also builds on controversial models of so-called modified gravity, such as the Modified Newtonian Dynamics (MOND) theory of Mordehai Milgrom.

Verlinde鈥檚 calculations fit the new study鈥檚 observations without resorting to free parameters 鈥� essentially values that can be tweaked at will to make theory and observation match. By contrast, says Brouwer, conventional dark matter models need four free parameters to be adjusted to explain the data.

鈥淭he dark matter model actually fits slightly better with the data than Verlinde鈥檚 prediction,鈥� says Brouwer. 鈥淏ut then if you mathematically factor in the fact that Verlinde鈥檚 prediction doesn鈥檛 have any free parameters, whereas the dark matter prediction does, then you find Verlinde鈥檚 model is actually performing slightly better.鈥�

Galactic lenses

Brouwer鈥檚 study takes advantage of catalogues of distant galaxies released in and and looks at regions close to the visible disc of each galaxy. These regions are where gravitational lensing should be bending light from distant galaxies beyond.

Using statistical algorithms that consider the shape and color of the background galaxies, the researchers inferred a lensing profile for the foreground galaxy. It鈥檚 a bit like projecting an image onto a warped and uneven sheet of glass and then, knowing what the original image looks like, figuring out the optical properties of the glass sheet from what we see on the far side.

From the gravitational distortions inferred for each foreground galaxy, the researchers devised a lensing profile based on Verlinde鈥檚 gravitational model, and another based on a conventional dark matter approach.

So if Verlinde鈥檚 is the better match, what鈥檚 the problem? Gravitational heresy. Verlinde鈥檚 gravity is stronger and dies off more slowly with distance compared with the models of Newton and Einstein.

To most physicists and astronomers today, that鈥檚 an issue, to put it mildly. Newton鈥檚 and Einstein鈥檚 theories of gravity have been so rigorously and comprehensively validated experimentally that it borders on sacrilege to suggest gravity could be something other than what they describe. String theorist Lubos Motl in a recent blog post: 鈥淚 wouldn鈥檛 okay this wrong piece of work as an undergraduate term paper.鈥�

Milgrom, however, supports the work. He also points out that according to his own 2013 , MOND produces similarly impressive results as Verlinde鈥檚 gravitational model does in Brouwer鈥檚 study.

鈥淢y equations work differently than Milgrom鈥檚, and in the case of [galaxy] clusters this can be quite important,鈥� Verlinde says. But in the case of Brouwer鈥檚 work, 鈥淭hey put in the formula I get,鈥� he says, 鈥渁nd I have to admit it鈥檚 the same formula that Milgrom would have got, and鈥� they just put it on the data. It looks like a fit.鈥�

Reference: