systemic

Astronomers worldwide staggered into work this morning, some of them rudely elbowing their way to the front of the lines at the espresso machines, clear evidence that events surrounding the January 2010 ’606 holiday season have finally drawn to a close.

Hopefully the data will turn out to be of high quality! As I mentioned in yesterday’s post, ground observers in both Europe and North America were out in force for the event, collecting photometric and spectroscopic data. The action was covered from space as well. We were awarded a generous 84-hour block of time on Warm Spitzer. The telescope started collecting 4.5-micron photometry more than a day prior to the secondary transit, and ended more than two days after the periastron passage.

What do we hope to learn? By observing the run-up to the secondary transit, we should be able to establish an improved baseline temperature for the planet, which should afford a better sense of how much tidal heating is occurring. And during the days following periastron, we expect to see a near-complete drop-off in flux from the planet as the periastron nightside hemisphere rotates fully into view. The 2007 observations came to a frustrating end just as this should have been starting to occur.

In addition to the secondary eclipse and the ground-based observations, Guillaume Hebrard and his collaborators were awarded 19 hours on Warm Spitzer to observe the primary transit at 4.5 microns. Their photometric time series will enable an improved radius measurement for the planet — both because of the highly accurate photometry and because the effects of stellar limb darkening are negligible in the infrared. Their time series will establish a very precise ephemeris for the transit, which will enable future observations to monitor the system for orbital precession.

Looking forward to the results…

It’s 4pm Wednesday Jan 13th here in Santa Cruz, and the HD 80606b transit has been underway for a few hours. A whole slew of observers worldwide are watching the event, with Northern Europe getting the best view (if the weather is clear).

Last weekend, the Spitzer telescope carried out an 84-hour observation of the system during the window surrounding the secondary eclipse. Our goal was to watch the planet heat up and then cool down rapidly as the unheated night side rotates into view.

Good luck to everyone who’s out there on the sky!

A quick addendum to the previous post. After a rather lengthy and undeserved “vacation”, Transitsearch.org is back on the air. The old website is running as a placeholder, and updated content will follow on soon.

I’ve moved the front-end of the transitsearch site to the hosting service that runs oklo.org, so the real URL is www.oklo.org/transitsearch/ By Dec. 10th, the domain name transfer will be complete, and the old www.transitsearch.org address should properly redirect.

Further updates can be had by subscribing to Transitsearch.org’s twitter stream: http://twitter.com/Transitsearch. We’re planning events to surround the next ’606 day, and we’re also planning to organize a campaign for the HAT-P-13c transit opportunity that’s centered on April 12, 2010.

Full-resolution Poster-sized .pdf of the above.

The next HD 80606 transit is coming up this week. While the sky position of the star will be much more favorable during the coming January event, observers across the US have an opportunity to get photometric measurements of the ingress early Thursday morning.

The transit begins just after 11 AM UT on Sept. 24, and will unfold over the next 12 hours, meaning that observers in Japan and East Asia will be able to catch the egress.

Josh Winn of MIT is organizing a repeat of the successful June campaign (detailed in this post). If you’re a capable photometric observer, and if you’re interested in participating in the campaign, definitely get in touch with him.

An unsung advantage of long-period transiting planets is that the occultations occur on a civilized timescale. An interval of 111.4357 days is long enough not to feel pressured, rushed, or in constant danger of getting scooped. This is in stark contrast, to, say, managing your affairs with a fixed 2.2185733 day turn-around time.

Earlier this summer, there were two papers, one by Pont et al. and one by Gillon which presented complete, leisurely analyses that combine all of the available photometric and RV data for the HD 80606 system taken through the Valentine’s Day 2009 transit. These papers adopted a fully Bayesian approach to analyzing the heterogeneous data sets, and were able to improve the system’s vital stats: The planet has a radius very similar to Jupiter. The full duration of the transit is close to 12 hours (and uncertain to a bit more than an hour). With high confidence, the planet’s orbit is badly misaligned with the stellar equator — just as expected from the Kozai migration hypothesis.

Last night, Josh Winn sent me a new preprint that reports results from an extensive campaign that he spearheaded to observe the June 4th/5th 2009 transit. June, to put it mildly, is not exactly an ideal time to observe HD 80606 from Earth. The nights in the Northern Hemisphere are short, and the star sets early. At any given spot, you can get at best a few hours of uninterrupted data. Nevertheless, it was of great interest to bag the transit. The ingress was weathered out during the February event, and so the analyses of Pont et al. and Gillon had to lean rather heavily on the Good Reverend Bayes.

Josh’s strategy was to recruit an East-to-West swath of observers in Massachusetts, New Jersey, Florida, Indiana, Texas, Arizona, California, and Hawaii. The idea was that 168 electoral votes would be enough to tilt the contest in favor of the good guys.

The multi-state strategy paid off. By stringing together the individual photometric blocks, the first half of the transit was nicely resolved. At the finish line, on the summit of Mauna Kea, the Keck telescope stepped up to the podium to obtain a series of mid-transit spectroscopic measurements that further confirmed the severe spin-orbit misalignment.

.ppt-ready higher resolution version

This is just the sort of project that underscores the great value of ad-hoc collaborations. The Florida ingress observations, for example, were made using the University of Florida’s recently refurbished Rosemary Hill Observatory, 30 miles from Gainesville. The DeKalb observations, made by Indiana amateur Donn Starkey, produced reduced data that were among the best in the entire aggregate. Mount Laguna Observatory, run by San Diego State University, has generated many cutting-edge exoplanet observations, including critical photometry in the Fall 2007 HD 17156b campaign. The University of Hawaii 2.2m telescope turned out photometry with astonishing rms=0.00031 precision. And as the cherry on top, the simultaneous commandeering of not one but two major telescopes on Mauna Kea? It seems that perhaps someone has made a Faustian bargain.

Image Source.

My colleague Garth Illingworth, who is well connected to the Space Telescope Science Institute, sent an e-mail to the UCSC Department this morning that details the ongoing repair and refurbishment of the Hubble Space Telescope.

HST repair day 4 EVA is ending. STIS repair done and aliveness test shows that it is working. Full functional needed to verify but early indications good. COS is looking good so hopefully we will have two uv spectrographs.

Tests on ACS during the crew sleep last night showed that the WFC camera is working and that it passed its initial functional tests – with preliminary results suggesting that read noise is possibly lower than before. The combination of ACS, WFC3 IR and WFC3 UV-Optical will make HST’s imaging capability the best ever.

The ACS HRC cannot be recovered due to the location of the short in the power path (location unknown before powering up ACS last night – so the hoped-for “back-powering” approach for HRC did not work out).

So we are 4 for 4 on instrument repair! A remarkable effort by the NASA GSFC/STScI folks who brought all this to fruition, along with the flight teams and the astronauts.

The bit of good news that really caught my eye was the apparently successful repair of the STIS imaging spectrograph. Before its failure in 2004, the STIS spectrograph (which can operate in both the visible and the ultraviolet) was used to make the iconic transit light curve of HD 209458, and to make the first measurements of the atmospheric contents of hot Jupiters.

In 2003, STIS was also employed to observe the transit of HD 209458b in the ultraviolet region of the spectrum surrounding the Lyman-alpha line of hydrogen (paper). The data suggest that the HD 209458 b transit has a depth of order 15% in Lyman alpha, indicating that a comet-like wind of hydrogen is flowing off the planet. The press releases surrounding this event produced perhaps the most dreadful artist’s impression in the entire exoplanetary canon, here’s a more restrained cartoon that shows the basic idea:

The deep Lyman-Alpha transit depth of HD 209458 has remained something of a mystery, and it will be very exciting to observe the transits of other planets in the UV. In particular, the results for progressively more eccentric planets such as X0-3, HAT-P-2b, HD 17156b, and HD 80606b should be very informative. If the irradiation varies drastically over the course of an orbit, how is the wind flow affected? It’s always nice when there’s an opportunity to set forth an eminently falsifiable prediction…

Jonathan Langton took the Spitzer 8-micron time-series for HD 80606b and transformed it into a movie of an actual extrasolar planet. The money-minded studio execs, having never seen the the successful prequel, decided that the full 30-hour version might not do well in theaters, so the original “Director’s Cut” had to be edited. The final result? Thirty hours of one-pixel, gray-scale footage have been compressed into a 10-second movie showing the excitement surrounding perihelion at a rate of 3 hours per second.

Be sure to watch for the secondary eclipse!

The visible universe contains of order 30,000,000,000,000,000,000,000 planets, and so this web log’s rather single-minded focus on HD 80606b (a staggering eight out of the nine most recent posts) is likely starting to wear a little thin, even for the Kid606 fan base. One more post, though, and then I’ll move along.

First, I was jazzed to get an e-mail from Mauro Barbieri (of 17156, etc. fame) reporting that two Italian amateur observers (Alessandro Marchini from Siena, Tuscany, and Giorgio Corfini, from Lucca, Tuscany) got discovery photometry of the HD 80606b transit on Feb. 13th/14th. Their light curves are of quite high quality, and, like all the European observations show the leisurely egress from transit:

Excellent work!

A few long-time readers may recall that in the transit fever post from several years ago, I tried on a “tough guy” persona with regards to partial transits:

The transit detection problem is tough in part because it’s extraordinarily easy for systematic effects to seemingly conspire to produce an apparent signal. I would not feel confident in announcing a transit until I’ve seen multiple full-transit light curves. On the other hand, though, the false alarms play an important role. They get observers out on the sky, and spur the collection of enough data to truly rule out an event.

This hard-line attitude resulted from catching numerous infections of ingressia in which a time-series seems to show a transit starting just as observations are ending:

and egressia in which a transit seems to be ending just as observations are starting:

With HD 80606b, however, it’s perfectly certain that we’re not dealing with a virulent case of egressia. The transit did occur and that it will occur in the future. This confidence stems both from the fact that there are at least seven independent photometric data sets showing the egress, and from the fact that the French-Swiss team (Moutou et al. 2009) observed the transit spectroscopically via the Rossiter-McLaughlin effect.

The Rossiter-McLaughlin effect arises when a transiting planet occults part of a rotating star. When a planet passes in front of the oncoming limb, it blocks out blue-shifted light, whereas it blocks out red-shifted light when covering the outgoing limb. The resulting distortions in the spectra are interpreted as a positive and then negative shift in the radial velocity of the star. The amplitude of this effect is thus due both to the spin velocity of the star as well as to the total flux blocked out during transit:

Moutou et al.’s detection of the Rossiter-McLaughlin effect for HD 80606b provided drop-dead confirmation of the transit, and also hinted that the planetary orbital plane is not aligned with the equator of the star (which is not surprising, given the probable history of the ’606 system). Here’s a re-working of the diagram from the Moutou et al. paper that takes the London and Arizona photometry into account (you may want to make your browser window wider):

Illustrator .ai file for above image

The Arizona and London photometry rule out transits longer than ~12 hours, which strengthens Moutou et al.’s conclusion that the system is far from having the stellar equator aligned with the orbital plane.

Earlier this week, I was having an e-mail conversation with Bruce Gary, who runs the Amateur Exoplanet Archive (a.k.a. AXA). The AXA is a repository for photometric transit data from small telescopes, and a first stop for anyone interested in the detection of planets via transit timing.

Bruce wrote:

By the way, does the Rossiter-McLaughlin effect refer to the Dean McLaughlin who speculated about Mars, and who worked at the Univ Michigan Observatory in the late 1950s & early 1960s?

A bit of ADS sleuthing reveals that the two McLaughlins are one and the same. In 1924, Richard Rossiter and Dean McLaughlin simultaneously published the first measurements of spin-orbit alignment in eclipsing binary systems. Both men were at the University of Michigan — Rossiter as an assistant professor and McLaughlin as a 23-year old graduate student. McLaughlin used the famous eclipsing binary Algol to measure the time-dependent radial velocity skew in the brighter star of the system during the partial eclipse. His paper, “Some Results from a Spectroscopic Study of the Algol System”, makes a nice read today, and has garnered 45 citations since 2000. Its single figure shows the now-familiar effect, albeit with a factor-of-a-thousand increase in the scale of the y-axis:

McLaughlin remained at the University of Michigan during a productive career that ended with his untimely death in 1965. He seemed to have had a sensibility that was quite in line with oklo.org. Consider, for instance, this abstract from 1944:

Bruce later wrote back with small-world anecdote:

As I was finishing high school my father counseled me to not choose astronomy for a profession because Dean McLaughlin’s two boys were in his Ann Arbor High School English class and their clothes gave the impression that the McLaughlins were a poor family! That influenced my decision to enter the University of Michigan’s School of Engineering, but after a year my childhood hobby won out and I switched to Literature, Science and Arts so I could major in astronomy.

Nice!

After 10 days of no news, definitively flat news (Arizona) and tantalizing hints in my inbox, the HD 80606b transit story is resolving itself dramatically.

Earlier today, Stephen Fossey, Ingo Waldmann and David Kipping submitted their paper on the detection. I based the diagram on the results of their photometry, which points to a twelve hour transit, and a planetary radius just larger than Jupiter:

The Fossey et al observations were made using two small telescopes at the University College London’s observatory in Mill Hill, North London. (Co-author Ingo Waldmann is a final-year undergraduate project student.) It’s certainly been a long time since an observational astronomical discovery of this magnitude has made from within the London City Limits!

Also in my inbox this morning was an e-mail from Jose Manuel Almenara Villa, who made the definitive initial observation of HD 17156 (and made the initial announcement on the comment section of this weblog). He writes, I know it’s late, but here there are the data from Tenerife. The egress is fully there, fully present. Nice work!

And then, no more than an hour ago, another dramatic update. In an e-mail to myself and Jean Schneider, Enrique Garcia-Melendo writes:

Dear Greg and Jean,

We observed the transit of HD80606b.

Please find attached the submitted paper to the ApJ. The manuscript will also appear at http://arXiv.org/abs/0902.4493

Best regards,
Enrique Garcia-Melendo

Title: Unconfirmed Detection of a Transit of HD 80606b
Authors: E. Garcia-Melendo and P. R. McCullough
Categories: astro-ph.EP
Comments: Submitted to ApJ, 11 pages, 4 figures.

We report a times series of B-band photometric observations initiated on the eve of Valentine’s day, February 14, 2009, at the anticipated time of a transit of the extrasolar planet HD 80606b. A transit model favored by the data has minimum light of 0.990 times the nominal brightness of HD 80606. The heliocentric Julian date (HJD) of the model’s minimum light is 2454876.33, which combined with the orbital period P = 111.4277 pm 0.0032 days, longitude of periastron, omega = 300.4977 pm 0.0045 degrees, and time of mid-secondary eclipse HJD 2454424.736 pm 0.003 (Laughlin et al. 2009), refines the eccentricity, e = 0.9337 +0.0012 -0.0004}, and the inclination, i = 89.26 +0.24 -0.04 degrees. The duration of the model transit is 0.47 days, and its four contacts occur at HJD 2454876 plus 0.10, 0.24, 0.42, and 0.57 days. We observed only the last two contacts, not the first two. We obtained “control” time series of HD 80606 on subsequent nights; as expected, the “controls” do not exhibit transit-like features. We caution that 1) the transit has not been confirmed independently [note: no longer true.]; 2) we did not observe the transit’s ingress; 3) consequently, we cannot reliably measure the relative sizes of the planet and its star in a model-independent manner, and 4) hence, the other values derived herein are also model dependent.

Now here’s the kicker — the Garcia-Melendo & McCullough paper was submitted on Feb. 23rd…

Update: I just heard from Shigeru Ida at Tokyo Institute of Technology, who has coordinated a number of photometric campaigns by amateur observers in Japan. It turns out that it was either rainy or totally cloudy on the night of the transit ingress (Feb. 13/14) for all of the observers. Bummer. The following night, the conditions were a little better, allowing several observers to get noisy baseline data.

I’m very pleased to be able to announce that HD 80606b is a transiting planet!

It looks like priority of discovery goes Claire Moutou and the French and Swiss team, who beat at least one other team to submission by a matter of hours. I’m attaching a draft of the French and Swiss Team’s paper that was just sent to me. Congratulations to Everyone involved!

Here’s the preprint.

Details to follow…