High proper motion stars mess with Google Sky

[tweetmeme only_single=false service=”wp.me” source=”allinthegutter”]

I was alerted to this by a post on the Bad Science forum. There one of the posters had used the position of Barnard’s Star to estimate the date a Google Sky image had been taken. Barnard’s Star has the highest proper motion of any star known. This means it moves across the sky compared to background stars. Looking at the image I wondered why Barnard’s star appeared only once as a single, very blue image. These colour images are made by combining images of the sky taken in different colour filters. These images are often taken years apart so for an object such as Barnard’s star which moves very fast across the sky the positions in each filter will be different. Hence you would expect to see one blue, one green and one red image in three different positions, this doesn’t appear to happen.

I decided to check out a few other high proper motion stars to see what they looked like. Proxima Cent is a bit weird, I can’t seem to identify it on the image, perhaps it is the blue thing on top of a background star, there is certainly a bit of noise where the UKST I plate (Google Sky uses a combination of data from the UKST, POSS telescope and the Sloan Digital Sky Survey plus a few other sources of more detailed images) position from the 1970s is. A better example is Kapteyn’s Star is a better example. Notice the bright very blue object in the upper right, that is the blue image from 1975, while the noisy thing in the middle is around the position of the red image from 1998. You can see a better subtraction for Luyten’s star.

Frankly I’m not sure about the finer points of how Google Sky make their colour images. This is clearly an artifact of the way in which they combine the images. Anybody able to use this to work out why this is happening?

A previously seen planet?

[tweetmeme only_single=false service=”wp.me” source=”allinthegutter”]

At the end of last year a couple of papers appeared with some very promising looking direct images of extrasolar planet candidates. Until now the bulk extrasolar planets (i.e. planets outside our solar system) have been found either by the radial velocity method where the motion of the parent star being pulled around by the planet is detected or the transit method where the planet obscures a portion of the parent star, blocks some of the light that would otherwise reach us here on Earth and makes the star appear a bit dimmer. One of the candidate direct images was of a planet around the nearby young star Beta Pictoris, the discovery paper by Lagrange and collaborators is here and here and the press release is here. These direct images are very difficult to acquire as the star is much, much brighter than the planet (in the case of Beta Pic about 1500 times brighter) and the atmosphere and telescope optics smear out the star’s light, covering the spot on the sky where the planet is. The group led by Lagrange used the Very Large Telescope in Chile along with the NaCo instrument (which both blocks out most of the light from the parent star and corrects for some of the atmospheric smearing). This has allowed them to image what looks like a planet near the star. Of course it could just be another, fainter, unrelated star behind Beta Pic, in these cases you need to come back a few years later to check the planet is moving through space along with the parent star to make sure. However the chance of this just being coincidence is pretty small.

So why am I writing about this now? Well a paper has appeared that may indicate this planet was detected before, in 1981. Back then Beta Pic was seen to dim briefly, as if a planet passed in front of it. This of course begs the question “was the imaged planetary candidate responsible for the transit?” This is the question the authors try to answer. A planet will only transit if you are looking at the system edge-on and we have a clue that the Beta Pic system is very close to edge-on. Like many young stars Beta Pic has a disk of material around it that is thought to form planets. We know that the disk around Beta Pic is pretty close to edge on and you’d expect the planets in any system to orbit roughly in the plane of the disk. Hence it is possible the planetary candidate could transit in-front of the star. The authors then go on to try to work out (assuming the planetary candidate and the transiting body are the same thing) when a transit would happen again and what the planet’s orbit is. They find the most likely solution is a planet orbiting Beta Pic at a distance of eight times the Earth-Sun distance every 16-19 years. Both the direct detection and the transit suggest the planet is a gas giant.

The idea that the transit of a planet across its parent star could have been detected in 1981 sort of shows that astronomy is a passive science. In most research you design an experiment, have complete control over it, carry it out and note down the result. In astronomy you can’t grab two bottles of chemicals off the shelf and mix them, you can only look. If there is a planet around Beta Pic in a 16-19 year orbit then it was also there in 1981, it was also transiting at the end of the last century and in the mid-60s, just nobody was looking. Almost everything we can study, measure and analyse in astronomy is already out there, we just haven’t looked hard enough yet.