Tau Bootis is the weirdo among the exoplanet stars.
It's the youngest of the set: 2 Gyr, whereas most are older than 5 Gyr. This is due to a selection bias because younger stars have bigger star spots which can ``ape'' the spectral signature of an exoplanet.
This has happened a couple of times already, but in each case follow-up photometry has shown that the variations aren't planet induced.
The radial velocity variations have a much larger amplitude than the other exoplanet stars, about 450 m/s.
The orbital period is about the same as the rotation period of the star. This is a problem because the star is under "active" for it's rotation. So, does the planet influence the star's orbit here (but not for any of the other 51 Peg-type planets???) or vice versa.
The period I *expect* from stars of Tau Boo's mass and age is 5 days, not 3.3 days. But I've been monitoring the rotation for 15 years and it's definitely shorter than 5 days. Weird, weird, weird.
Tau Bootis has a weird other period that doesn't "fit" with anything I can think of. I'm still writing a paper on this, but the punch line of "duhhhhh I don't know" isn't very interesting:-)
Interested parties can get our papers at http://cfa-www.harvard.edu/~donahue/P reprints/. In the newest paper (I'll put it up on the page tomorrow, PROMISE), we definitely show that there's NO transit of the planet across the stellar disk.
The St. Andrews group are REALLY good people. And, I"m saying that not just because they said nice things about my research recently!:-)
Well, as for Venus, I don't know of any papers doing this in terms of spectroscopy, but there is a very famous example of how you can see Venus' atmosphere and show that it is VERY dense:
During transits of Venus
just as the ingress phase is ending, you see a "tab" of darkness extending from the edge of the solar disk to Venus' disk that's caused by the dense atmosphere.
Coincidentally, there is a transit of Mercury tomorrow which can be seen from the West Coast. Transits of Venus are much more infrequent, but have the weird property that two occur within a few years of each other, then not again for something like 200 years. The next two are in 2004 and (I think) 2012, so lucky us!:-)
-----------
Paul Butler had an interesting comment in the newspaper article suggesting that it would be possible to detect the planet's atmosphere during a transit.
What I suspect he's referring to would be the appearence of extra lines in the star's spectrum during a transit caused by absorption of starlight by the planet's atmosphere. The problem with this is that there isn't much light being absorbed, so the effect will be incredibly tiny. But with a LOT of photons it might show up. You could co-add all the in-transit spectra you get over time.
Heh. Having said this, it occurs to me that the widths of the spectral features of the star are much wider than the variations in wavelength shift that are used to detect the planet, but I don't have a clue what the widths of these "overlayed" features would be!
BUT if it's an atmosphere and it's cold (compared to the star's photospheric temperature, which even at the small orbit is completely reasonable) molecular bands would be where I'd look. They'd be broad because of all the many rotational and vibrational lines.
It's still a long shot, but three years ago, so was the entire idea of detecting exoplanets this way!
And, the chances of this working (IMHO) are a lot higher than the people who are looking for variations in the photometric light curve due to reflected stellar light from the close-in planet. (But the information gathered from this star will provide a better estimate of the expected flux amplitude --- I haven't tried to calculate it yet.)
Work is going to be VERY INTERESTING tomorrow and Tuesday!
The star is inactive compared to the Sun. I can't give specifics (yet) --- I only got a copy of MY data for this star on Friday, but I can tell from the raw data that it's probably at the level the Sun was at during the Maunder Minimum.
However, sunspots can mimic the RV variations if they are large enough. We looked into this problem in our first paper on the subject (Henry et al. 1997, ApJ, 474, 503), and in more detail in another paper (Saar & Donahue 1997, ApJ, 485, 319).
And there is a case where this happened!
The star HD 166435 is young, has a 3.8-day rotation, and the motion of the spots across the surface cause enough photocenter wander to influence the radial velocity variations that it mimic the effects of an exoplanet. We (Henry with precision photometry, and the HK Project at Mount Wilson with spectrophotometry) were able to show (before the paper was submitted for publication) that the variations coincided with variations in the photosphere and the chromosphere of the star with the same period.
The chromospheric variation is the exoplanet killer since the flux variations that are observed have to come from stellar activity.
Needless to say we're trying very hard to keep up with all exoplanet announcements, and even rumours to get a handle on the properties of the stars they orbit.
Actually, the transit that was reported was discovered by Greg Henry, of TSU using telescopes at Fairborn Observatory (south of Tuscon, AZ).
These telescopes do the most precise photometry ever achieved, working to about 0.001 magnitudes on a night-to-night basis, and about 0.0002 mags for long-term variations. That's ALMOST good enough to montior irradiance changes for stars that vary as little as the Sun does. On a very good night, with lots of overlapping data, these telescopes could almost detect a transit of an Earth-sized planet.
There are two published papers on using these telescopes to look for transits in exoplanet systems. A third has been accepted for publication by the Astrophysical Journal and will come out in the March 10, 2000 issue. (I'm one of the authors.) Preprints of the papers are all available on one of my webpages:
(I'll get the preprint of the 3rd paper up there on Monday.)
It's great to see that a transit has finally been observed! We were starting to get worried... The search for transits is being done in collaboration with a long-term program to better understand the stars they orbit also done at Fairborn and with Mount Wilson's HK Project.
Given the *current* state of affairs, which I would list as:
IE5 has been out for several months
IE5 has better compliance with the DOM-1 than NN4
Mozilla is Waaaaaaay behind in getting out
IT drones are starting to abandon NN as browser to support, either in using it as a client OR supporting it in their internal and external webpages
Then, if I were M$, I would doing these things:
Working on a new version of IE to come out RIGHT after NN5 with lots of new goodies (it doesn't matter if they're in the DOM standard or not) in order to push the bar away from NN5.
Hold back on porting IE to non-M$ platforms as much as possible. If I HAD to get something out, make it bug-ridden, and have it interfere with other things on that/those platform(s) as much as possible in order to nudge people towards NT/W2K.
Encourage IT people to support IE's unique features as much as possible, and discourage them from holding back on their own site development for NN5 to catch up. Label NN as "obsolete" and make a point of creating as many sites that only work on IE.
Hope like hell that the Mozilla people drop the ball and NN5.0 is buggy, non-uniform in terms of cross-platform performance, and is as unstable as IE, esp. on NT/W2K.
It's straightforward to see how this impacts the Linux community. First, we become stuck with an "outdated" browser on an operating system that isn't allowed to have IE. At the same time, all of the effort is pushing towards catching up to IE, which makes it harder to get 3rd party support for non-M$ OS's, and diminishes our ability to use the Internet.
What can be done? At this point, I'm just not sure. I spent this AM screaming at Netscape because every platform renders PostScript differently - so you get different printed output depending on *which* Netscape you use! (I expected some differences but it was as much as what you'd expect from N different browsers!)
So much for "cross-platform". I keep hearing that Mozilla will have true cross-platform support, but at this point, I have little faith in this claim. If it isn't, we can expect to have a VERY buggy release of NN5.0, which will pretty much bring down the axe (see #4 at top). "Stop Rewarding Stupidity" Bob
Interesting --- I tried to complain about the lack of QT4 support for Linux, but the e-mail addresses that were supplied at the QT website were not valid.
Has anyone had better luck? As near as I could see, they weren't interested in any OS that wasn't MacOS or Win9X/NT.
This to me is the chief problem, and is why I switched from BSDI to Linux: 3rd party support. I've written letters, called companies, even had companies call companies, but the ONLY response we ever get is "well, just buy a Win9X/NT box!".
This whole "save Mozilla" campaign is futile. It's already far too late. Netscape was once ahead of IE because they had the edge in that M$ took forever to get onto the Internet. The only way to regain that edge is to come up with something as important as the WWW in collaboration with everyone BUT M$, and leave them out in the cold. However, the WWW is no longer a realistic venue for this because there's little that can be done that won't just be "the minority" acting in the face of a (perceived or actual, it doesn't matter) majority using M$/IE.
From this point forward, if M$/IE doesn't have it, it won't catch on - at least until M$/IE supports it, and if only M$/IE has it, everyone else will just be told to get IE.
The war (over the WWW) was over 2-3 years ago. When IE4 beat NN4 in release date. IE5 has now beaten NN5 by *months* (really, *years*) and all the hopes about Mozilla isn't going to regain that time lost. There's nothing *in* NN5 to give it dominance over IE5 --- worse, without widespread plug-in support (even widely used plugins like QT4, etc.) it's dead before it's even released.
I've no intention of using IE myself, but no company is going to care about that. Unless that changes, it doesn't matter anymore what great things Mozilla might be able to do someday when it gets released.
Slashdot Mirror
This has happened a couple of times already, but in each case follow-up photometry has shown that the variations aren't planet induced.
The period I *expect* from stars of Tau Boo's mass and age is 5 days, not 3.3 days. But I've been monitoring the rotation for 15 years and it's definitely shorter than 5 days. Weird, weird, weird.
Interested parties can get our papers at http://cfa-www.harvard.edu/~donahue/P reprints/. In the newest paper (I'll put it up on the page tomorrow, PROMISE), we definitely show that there's NO transit of the planet across the stellar disk.
The St. Andrews group are REALLY good people. And, I"m saying that not just because they said nice things about my research recently! :-)
Coincidentally, there is a transit of Mercury tomorrow which can be seen from the West Coast. Transits of Venus are much more infrequent, but have the weird property that two occur within a few years of each other, then not again for something like 200 years. The next two are in 2004 and (I think) 2012, so lucky us! :-)
-----------
Paul Butler had an interesting comment in the newspaper article suggesting that it would be possible to detect the planet's atmosphere during a transit.
What I suspect he's referring to would be the appearence of extra lines in the star's spectrum during a transit caused by absorption of starlight by the planet's atmosphere. The problem with this is that there isn't much light being absorbed, so the effect will be incredibly tiny. But with a LOT of photons it might show up. You could co-add all the in-transit spectra you get over time.
Heh. Having said this, it occurs to me that the widths of the spectral features of the star are much wider than the variations in wavelength shift that are used to detect the planet, but I don't have a clue what the widths of these "overlayed" features would be!
BUT if it's an atmosphere and it's cold (compared to the star's photospheric temperature, which even at the small orbit is completely reasonable) molecular bands would be where I'd look. They'd be broad because of all the many rotational and vibrational lines.
It's still a long shot, but three years ago, so was the entire idea of detecting exoplanets this way!
And, the chances of this working (IMHO) are a lot higher than the people who are looking for variations in the photometric light curve due to reflected stellar light from the close-in planet. (But the information gathered from this star will provide a better estimate of the expected flux amplitude --- I haven't tried to calculate it yet.)
Work is going to be VERY INTERESTING tomorrow and Tuesday!
Bob Donahue
The star is inactive compared to the Sun. I can't give specifics (yet) --- I only got a copy of MY data for this star on Friday, but I can tell from the raw data that it's probably at the level the Sun was at during the Maunder Minimum.
However, sunspots can mimic the RV variations if they are large enough. We looked into this problem in our first paper on the subject (Henry et al. 1997, ApJ, 474, 503), and in more detail in another paper (Saar & Donahue 1997, ApJ, 485, 319).
And there is a case where this happened!
The star HD 166435 is young, has a 3.8-day rotation, and the motion of the spots across the surface cause enough photocenter wander to influence the radial velocity variations that it mimic the effects of an exoplanet. We (Henry with precision photometry, and the HK Project at Mount Wilson with spectrophotometry) were able to show (before the paper was submitted for publication) that the variations coincided with variations in the photosphere and the chromosphere of the star with the same period.
The chromospheric variation is the exoplanet killer since the flux variations that are observed have to come from stellar activity.
Needless to say we're trying very hard to keep up with all exoplanet announcements, and even rumours to get a handle on the properties of the stars they orbit.
Bob Donahue
These telescopes do the most precise photometry ever achieved, working to about 0.001 magnitudes on a night-to-night basis, and about 0.0002 mags for long-term variations. That's ALMOST good enough to montior irradiance changes for stars that vary as little as the Sun does. On a very good night, with lots of overlapping data, these telescopes could almost detect a transit of an Earth-sized planet.
There are two published papers on using these telescopes to look for transits in exoplanet systems. A third has been accepted for publication by the Astrophysical Journal and will come out in the March 10, 2000 issue. (I'm one of the authors.) Preprints of the papers are all available on one of my webpages:
(I'll get the preprint of the 3rd paper up there on Monday.)
It's great to see that a transit has finally been observed! We were starting to get worried... The search for transits is being done in collaboration with a long-term program to better understand the stars they orbit also done at Fairborn and with Mount Wilson's HK Project.
Bob Donahue
Given the *current* state of affairs, which I would list as:
Then, if I were M$, I would doing these things:
It's straightforward to see how this impacts the Linux community. First, we become stuck with an "outdated" browser on an operating system that isn't allowed to have IE. At the same time, all of the effort is pushing towards catching up to IE, which makes it harder to get 3rd party support for non-M$ OS's, and diminishes our ability to use the Internet.
What can be done? At this point, I'm just not sure. I spent this AM screaming at Netscape because every platform renders PostScript differently - so you get different printed output depending on *which* Netscape you use! (I expected some differences but it was as much as what you'd expect from N different browsers!)
So much for "cross-platform". I keep hearing that Mozilla will have true cross-platform support, but at this point, I have little faith in this claim. If it isn't, we can expect to have a VERY buggy release of NN5.0, which will pretty much bring down the axe (see #4 at top). "Stop Rewarding Stupidity" Bob
Has anyone had better luck? As near as I could see, they weren't interested in any OS that wasn't MacOS or Win9X/NT.
This to me is the chief problem, and is why I switched from BSDI to Linux: 3rd party support. I've written letters, called companies, even had companies call companies, but the ONLY response we ever get is "well, just buy a Win9X/NT box!".
This whole "save Mozilla" campaign is futile. It's already far too late. Netscape was once ahead of IE because they had the edge in that M$ took forever to get onto the Internet. The only way to regain that edge is to come up with something as important as the WWW in collaboration with everyone BUT M$, and leave them out in the cold. However, the WWW is no longer a realistic venue for this because there's little that can be done that won't just be "the minority" acting in the face of a (perceived or actual, it doesn't matter) majority using M$/IE.
From this point forward, if M$/IE doesn't have it, it won't catch on - at least until M$/IE supports it, and if only M$/IE has it, everyone else will just be told to get IE.
The war (over the WWW) was over 2-3 years ago. When IE4 beat NN4 in release date. IE5 has now beaten NN5 by *months* (really, *years*) and all the hopes about Mozilla isn't going to regain that time lost. There's nothing *in* NN5 to give it dominance over IE5 --- worse, without widespread plug-in support (even widely used plugins like QT4, etc.) it's dead before it's even released.
I've no intention of using IE myself, but no company is going to care about that. Unless that changes, it doesn't matter anymore what great things Mozilla might be able to do someday when it gets released.