Printer version

Nulling Starlight

Posted on Wednesday 26 November 2003 to Quintessence

There are then innumerable suns, and an infinite number of earths revolve around those suns, just as the seven we can observe revolve around this sun which is close to us.

[Giordano Bruno, 1584]

Giordano Bruno was the first to envisage planetary systems in revolution around distant stars but it took more than four hundred years before his speculation was proven to be correct. In 1995, Geneva-based astronomers, Michel Mayor and Didier Queloz identified a rythmic shift in the light spectrum coming from a star known as 51 Pegasi which was highly suggestive of the star being tugged at by an invisible body with about half the mass of the planet Jupiter. This behaviour of 51 Pegasi was confirmed shortly after by Geoff Marcy and Paul Butler who had themselves been searching for extrasolar planets since 1987. The door was opened and since then more than a hundred extrasolar planets have been found using this technique, by observing the periodicity of the doppler shifts in the light coming from the star it is possible to ascertain the masses and distances of these planetary bodies. But the technique is really only sensitive to planets within the same order of magnitude as the planet Jupiter and the evidence is indirect, so far no one has been able to directly observe an extrasolar planet.

Stars are more than a billion times brighter than planets and so the main difficulty with directly viewing extrasolar planets is that their feebly reflected light is completely swamped by the dazzling glare of their parent star.

Also the most interesting planets to look for are ones that are similar to earth and these are most likely to be very close to the star. Viewing our own solar system from a distance of 1 parsec, the planet earth would be a mere one second of arc away from the sun. Making a bad situation worse, one would have to look further than 1 parsec to find enough candidate systems to study, at least 10 parsecs away and preferably 25 parsecs.

It is, however, possible to reduce the brightness ratio between the star and the planet to about a million to one by restricting observations to the infrared band and, as luck would have it, this band (from 6 to 18 痠) also happens to be one of the most useful since it also contains the spectral lines of many of the most interesting molecules. These are water, ozone, carbon dioxide, all useful indicators for the presence of life (at least as we know it). Unfortunately, the spectral lines for oxygen are outside of this range, but the presence of ozone in the atmosphere is a strong indicator for the presence of oxygen as well.

But even with this restriction, a range of a million to one is still far too great for direct planetary observation so it has been proposed that one way to go about it is to suppress the starlight by cancelling it with itself, a technique which is known as nulling interferometry.

Schematic diagram of a standard stellar interferometer and the resultant fringe pattern at the two beam-splitter outputs. Conceptually the goal is to place the star at the bottom of a deep destructive fringe, and an accompanying terrestrial planet (TP) near the top of a constructive fringe.

The basic premise of nulling interferometry is conceptually quite simple: combine the starlight that arrives at a pair of telescopes so that at the centre of the image the two waves are exactly 180 degrees out of phase and effectively makes the star disappear. In the language of interferometry, a deep destructive fringe is to be placed across the star but light coming from sources that are offset slightly from it (i.e. from planets) gets added rather that substracted so that even though the star is completely blanked out, planets at the right locations (near a constructive fringe) are not attenuated.

Nulling interometry can be performed from earth, especially when imaging Jupiter sized planets, but the best place for viewing earth-sized planets close to their suns is from a space-based platform. Both NASA and and ESA have plans to launch infrared nulling interferometry telescopes into space over the next couple of decades. The ESA offering, known as DARWIN will consist of a "flotilla" of six individual spacecraft flying in formation while NASA's Terrestrial Planet Finder is based on four.

DARWIN

Terrestrial Planet Finder

Images of planetary systems are not formed by direct imaging but are reconstructed after measurements have been made with the array of telescopes in multiple configurations. The images is built up by rotating the array in a plane perpendicular to the line of sight to the star with measurements made at successive position angles. The data consist of a time series of the planet?s signal as it rises and falls through the fringe pattern of the rotating array.

By sampling using different spacings between the telescopes and at diffeent wavelengths it is possible to reconstruct an image of increasing detail. The reconstruction process does, however, add its own artifacts, note in the following (simulated) image the planetary doppleganger and the aliased planet at twice the orbital distance. These can be overcome with more sophisticated sampling techniques.