Scattered starlight may soon reveal the presence of extrasolar planets that cannot be detected by any other means, according to a pair of scientists in India. But some other experts say the method is best suited to studying the properties of known exoplanets ย– not turning up new discoveries.

Astronomers have already discovered about 155 extrasolar planets by watching how they make their host stars wobble or dim as they circle around them. But these methods are best suited to detecting so-called "hot Jupiters" ย– giant planets that orbit close to their stars, leaving any smaller or more distant planets unseen.

Now, Sujan Sengupta and Malay Maiti of the Indian Institute of Astrophysics in Bangalore say astronomers may soon be able to detect these elusive worlds too ย– providing they have atmospheres. Light bouncing off particles in their atmospheres should become linearly polarised, with its electric field aligned in one plane, they say.

Other researchers have modelled the polarisation signal expected from large, spherical planets travelling in tight circles around their host stars. But Sengupta and Maiti have now expanded these calculations to include planets that are slightly squashed at their poles because of their own spin ย– like Jupiter ย– as well as planets that are travelling in elliptical orbits around their host stars, which are less likely to make their stars wobble or blink by passing in front of them.Size isnย’t important

They found that non-spherical planets should actually produce a stronger signal because the polarisation averaged over the surface of a perfectly round object tends to cancel itself out.

Because the polarisation signal depends strongly on the angle between the star and planet, the researchers also argue that detecting any periodic change in polarisation implies the presence of a planet.

"It's not the amount of polarisation but the systematic time variation that will be important in detecting exoplanets," Sengupta told New Scientist. He says the method should work regardless of the planet's size, mass and distance from the star ย– and even with relatively imprecise polarisation measurements.

"According to my proposal, polarisation can detect exoplanets even if they cannot be detected by any other method," he says, adding that it should even pick up signs from Earth-like planets.Doubts over detection

But other researchers say the method may not be as robust as Sengupta believes. "I think there's a good chance to make a detection, but it's only the hot Jupiters that are going to be detected easily," says Sara Seager of the Carnegie Institution of Washington in Washington DC, US. She says the nearness of these large planets to their host stars means they will scatter more photons of starlight.

She also cautions starlight is likely to scatter multiple times inside a planet's atmosphere, potentially "washing out" the polarisation signal. "But beyond making a detection, polarisation is a powerful method for determining what is in a planet's atmosphere," she says, adding that such measurements of Venus from Earth revealed it to have sulphuric acid clouds.

Jim Hough, an astronomer at the University of Hertfordshire, UK, agrees. "I think the method is very important not for finding planets but for determining the inclination of their orbit, radius, surface reflectivity and properties of their atmospheres," he told New Scientist.

He has been searching for the polarisation of a known hot Jupiter around the star Tau Bootis but has so far not been able to find it. The planet is probably so massive and so close to the star that it is "stirring up the stellar surface and producing additional polarisation," he says.