?This Kind of Planet Had Never Been Seen Before?

Can you introduce the SWEEPS program to us?

This is a program to look for transiting extrasolar planets around stars in the galaxy bulge.  So far, all the planet detections, more than 200 or so, have been detected in nearby parts of the galaxy, nearby in the sense of 6,000 light years.  Here, we had a program to look all the way into the galactic bulge to see if there are planets just as there are in the solar neighborhood.  We used the Hubble space telescope to monitor one region in the sky, one field which has more than 180,000 stars, and we monitored this continuously for 7 days.  When a planet orbits around a star, if it passes in front of the star, then it would block a small amount of light in a very characteristic way.  By looking at this slight diminution of the light, we can detect the planet.  As the planet goes around the star, it will block the light and it will happen every time the planet crosses the star.  By monitoring for 7 continuous days, we can detect planets which have periods that are less than, roughly about 4 days (we need to see the planet at least twice during this period of 7 days).  We monitored 180,000 stars by looking at these transit signals, these small dimunitions of light, and we detected 16 planet candidates.

5 of these planets orbit their stars in less than an Earth day.  Is this a new kind of planet?

In some sense yes, because this kind of planet had never been seen before.  The lowest orbital period, the least amount of time in which a planet orbits around the star that was known before was 1.2 days.  This was slightly surprising that we found 5 of them orbiting around the star in less than one day, so in that sense, it is completely new.  But when you think about it, the program goes very deep—the Hubble telescope allows us to look at planets around extremely faint stars, extremely low mass, extremely low luminosity stars—and if you look at them, then yes, indeed, you would expect that the planet would be able to come closer to the star.  All the planet discoveries had not only been close by, but also the planet detection so far has been around slightly high mass stars.  Higher mass stars have high luminosity, so if they come close by, if they come too close, they can get evaporated or destroyed.  For example, if the sun had a planet with a period of less than one day, it would simply evaporate the planet.  But if you go and look for planets around low mass stars, these low mass stars are cool and have not so much radiation as the sun, so if the planet comes even closer by, it can survive.  That’s the reason why we found close-in planets.

Are you sure they are all genuine planets?  Can they be binary companion stars?

No, we are not sure that  all are planets.  In fact, we said that we are sure that about 50% of them are planets.  Sometimes these binary stars can mimic a planetary type transit; what happens is if the star is really small, then the radius of the star can be as small as the planet itself, then it can mimic a planetary type of signal. 

We not only detect planets but we also detect stars, eclipsing binaries and such. So if there is an eclipsing binary, when the star comes in front, then it blocks much more light than a planet does.  So we have detected 180 of such stellar binaries. We can use those to estimate what is the maximum number of such confusing cases there would be.  We know the mass distribution of these stars, so by looking at these stellar eclipses we can accurately estimate how many times we might get confused by these low mass stars.  So the number of planets that we have detected is way more than what can be caused by low mass stars masquerading as planets. 

We are not absolutely sure that all are planets, but that’s why we call them planet “candidates.”  But at least about half of them must be planets, and that is born out actually by our real measurements of mass in two cases. What really tells you it must be a planet is to be able to determine the mass of this transiting body through what is called the “wobble technique”, radial velocity observations.  In our case, most of our stars are so faint that you can’t really do these radial velocity observations even with the largest telescope available from the ground.  There were 2, actually, for which we did these radial velocity observations. We went to use this very large telescope in Mount Paranal in Chile, and we observed those 2 and in both those cases, they indeed came out to be planets.  That gives us a lot of confidence that our estimates could not have been wrong. 

Are your findings promising for the upcoming projects COROT and Kepler?

This will be actually a nice proof of concept for COROT and Kepler.  We have shown how you can detect planets. We have laid out the foundation for what kind of statistical estimates you can do, how you can do it, so this is a nice proof of concept for COROT and Kepler. 

Thank you Kailash Sahu

[1] Sahu K. C. et al, Nature, 443, 534 (5 October 2006)

Kailash Sahu works at the Space Telescope Science Institute (Baltimore, USA)

Interview by Thanh Tam Candice Vu