SuperWASP: Search for Extrasolar Planets

The name astronomical unit is shockingly narrow minded; it refers to our Solar System as though it is of Universal significance!

In fact it has been known for over a decade that there are other planetary systems. Most have been found from Doppler (i.e. speed) measurements of the wobble of the central star ("Sun") caused by the orbiting of the planets around it.

The 100 or so systems that have been recognised are mostly of large planets close to their parent star and are all among the closer stars (because they are easier to measure), and so are certainly only a small fraction of the total in our Galaxy. These are difficult measurements, definitely the domain of the professional observatory. However there is another method of seeking "exoplanets" being tried, which extends to amateur astronomer territory.

The transit of Venus will block about 0.1% of the Sun's light, from those parts of the Earth where it can be observed. This fraction is just the square of the ratio of angular diameters of Venus and the Sun at transit time - 60 arcsec and 30 arc minutes. In other planetary systems there will obviously be similar transits, to be seen from planets further out in the system, causing similar slight drops in luminosity.

In a few of these cases, it will happen by chance that the (very distant) Earth is also near the plane of the planetary disc. Of course, there will be no chance of seeing the line of the transit, because distant stars are seen as just a point, not a disc, but the dip in luminosity will be sufficient to witness.

Bigger planets, such as those already detected by orbital wobbles, will give bigger dips, but the real prize, for obvious reasons, is to discover Earth-size planets, especially those at about an AU from the parent star, and therefore with a temperature which allows liquid water. Do the 0.1% dips provide a method?

Those amateur astronomers that take a keen interest in variable stars (or rotating asteroids) already do better than 1%, and professional equipment gets down to 0.1%. But which stars should be measured? At this stage there is little guide as to which stars are most likely to have planets and in any case only a small fraction will be pointing at, or close to, the Earth.

Is it ridiculous to try to measure a million stars? Up to a few years ago it would have meant having thousands of independent photometers, and a thousand astronomers! Now, many people carry a million photometers in their pocket in the form of the CCD of a digital camera. A modest telescope or long camera lens can record thousands of stars in a single field of view. In a frame taken with a digital camera, each star is recorded as the charge of a number of electrons on a pixel (or small group of pixels).

To get an accuracy of 0.1% means one has to record a million or more electrons, but that is possible if the CCD has large pixels. Having a million or more pixels on a frame means data on many thousands of stars can be recorded in the digital file of the photograph. However, many stars are variable at this level - if you choose at random a target star and 3 or 4 others as reference stars, the chances are one of them will be variable.

Only about 1 in 1000 is expected to show transit-like behaviour. What makes the method viable is that the form of the dip for a transit - exactly symmetrical, with an initial drop lasting only a few minutes, then flat for a much longer time, then an equally rapid rise - is unlike the light curve of any intrinsically variable star, and so can be singled out.

Each star must of course be recorded repeatedly during a night in order to recognise this shape. It will be "winking at us to show that it has planets" in the words of one of the team, Carole Haswell. The data on the other 999 stars will not be wasted, it will be applicable to other problems as well as to exoplanets.

A prototype of an instrument called superWASP based on these ideas is operating (2004) in La Palma (www.superwasp.org - click on "Publications"), so there should soon be some results to look at. However, it's more than a compact digital camera, for various reasons.

First, the CCDs used in compact cameras are not the same as astronomical cameras, which need larger and preferably more pixels, of a type which store charge with very little loss - which makes them more expensive. Secondly, the lenses in compact cameras are tiny, whereas the La Palma instrument uses a much bigger lens to gather more light. It is a standard camera lens, but of high quality (200mm f/1.8). Amateurs will be able to approach, though not equal, these specifications. In fact amateurs have already recorded transits of a known exoplanet photometrically (see Tim Castellano and colleague's article in Sky and Telescope, March 2004, p 77-81, star HD 209458b, with a very clear 2% dip corresponding to Jupiter size).

Incidentally, a planet with rings (such as Saturn, Uranus or Neptune) is calculated to give an interestingly different profile (David Tytell, Sky and Telescope Jan 2004 p22). As a pointer to the future, the first CCD camera with a 35mm (i.e large) chip intended for the amateur market has just become available (see Sky and Telescope July 2004, p 96-102 ) from a company that specialises in this area.

The point of WASP is to record very large numbers of stars with relatively simple equipment and it is this scaling up which is not (yet) regarded as possible to amateurs. Why? Because the recognition of transit-like features in the many thousands of gigabytes of data has to be largely automatic, requiring large programs to be "trained" and rigorously standardised. However amateurs are currently making both photometric and astrometric measurements which five years ago would have been thought strictly professional, so let's hope the gap continues to narrow.

This image of Comet C/2001 Q4 (NEAT) was taken by SuperWASP on15th May 2004. The head of the comet is surroundedby the brightest stars in the constellation of Cancer.The comets' tail extends over 11 degrees across the sky: thisis 20 times larger than the full moon, and means a fist held atarms' length would only just cover the coment's tail.SuperWASP will revolutionise studies of extremely extendedobjects like this as a byproduct of its search for planets aroundbright stars.

Some technical details:

WASP stands for Wide Angle Search for Planets. Several members of the Open University Department of Physics and Astronomy, along with people from several other universities, are deeply involved inSuperWASP. It is called wide-angle because a 200mm lens is wide in an astronomical context, though not for photographers. The original WASP used one lens, SuperWASP currently uses five, each covering 8 x 8 degrees. The CCD's are large (2K x 2K 13.5 micrometre pixels, thinned and cooled). The instrument aims to record stars in the range 7 to 13 mag.

Update

Five years on from this article's original publication, Andy Norton brings us up to date on progress.