Posts Tagged ‘transits’


February 4th, 2009 4 comments

From a cell phone picture transmitted by an agent

The photo above is grainy, but what’s truly remarkable is that the depth of the dip is only 0.03%. Earth transiting in front of the Sun as seen from afar blocks roughly 0.01% of the Sun’s light. Look at the signal-to-noise of the bottom composite-average curve.

I can sure empathize with the CoRoT team. Their symposium date was set up long ago. Kepler is launching in a few weeks. The results of the Doppler surveys are suggesting that super-Earths with orbital periods of 50 days or less (with correspondingly high transit probabilities) are present around 30% of solar-type stars. Ground-based photometry is pushing below 0.5 millimagnitudes at 1-minute cadence. The pressure is on. And there’s an absolutely fascinating candidate planet that isn’t quite yet out of the oven, due to a paucity of high-precision radial velocities that would pin down the mass. What do you do?

I agree! You go ahead and announce.

Everything about CoRoT-7b reemphasizes the fact that planets are wont to turn up in every corner of parameter space to which observations are sensitive. In this case, a V=11 K0V star in the direction of the galactic anti-center displays 176 individual 1.5-hour 0.3 mmag photometric dips with a strict 0.854 day periodicity. These measurements suggest a 1.7 Earth-radii planet with a 20-hour year — a world that makes 51 Peg b look like Fargo North Dakota.

The abstract for Daniel Rouan’s talk at the meeting (transcribed from the cell phone photograph) describes the procedures that the CoRoT team has implemented to rule out the various false positives that can plague transit surveys. This gives a sense of the amount of follow-up work that needs to be done in order to secure a planet as small as this one (also, see comments section for this post, for many additional details):

To qualify/falsify the interpretation of the observed transits, we have considered different alternative interpretations: (1) transit of a main sequence star in front of a giant star — rejected by the measured log(g) of the target; (2) a grazing eclipse by a stellar companion — rejected by the radial velocity measurements (3) a weak Background Eclipsing Binary that would be inside the target mask — partially rejected by on/off transit photometric observations performed from the ground at angular distance from the target larger than 2 arcsec, and by high-resolution imaging at distances larger than 0.3 arcsec. (4) a triple system made of the target star (K0V) and a faint star (M5V) eclipsed by a giant planet or a dark stellar companion — rejected by the study of the transit colours which are the same as those of the main target. is quoting a significantly uncertain mass of 0.035 Jupiter masses (11 Mearth) for the planet, a figure that could have been arrived at via assumptions about the density and/or limits on the radial velocity detection. An 11 Earth-mass planet would induce an eminently detectable K=8 m/s RV signal, so it’s a bit odd that a firmer estimate of the mass isn’t available yet. The CoRoT “galactic anticenter” field is located in Monoceros, at RA~06h 45m, DEC~+0d, meaning that the candidate star is currently visible to HARPS (at air mass <2) all through the first half of the night. also states an age of 1.1 Gyr for the star, so youth, with its attendant stellar activity, could possibly be making it tough to get good velocity precision.

In any case, it’s a remarkable detection, and will be hugely influential as soon as the mass is confirmed. The planet is orbiting at only four stellar radii — with the star filling nearly a thousand square degrees of sky…

Categories: worlds Tags: , ,

The Big Swing

January 29th, 2009 3 comments

Image from computer modeling by J. Langton and D. Kasen.

HD 80606b — everyone’s second-favorite planet — is in the news! Our article describing the Spitzer Space Telescope’s 8-micron observations of the planet’s periastron passage made the cover of this week’s issue of Nature, and JPL has issued a press release on the results.

The planet has been a long-running topic here at, with the storyline developing over a series of posts during the past few years. A incomplete list might include:

Post one (older), two, three, four, five, though six (newer).

The outsize eccentricity of HD 80606b’s orbit leads to very brief, very intense encounters every 111.4 days as the planet swings through periastron. On the Nov. 20, 2007 encounter, we used Spitzer to monitor the 8-micron emission of the star and planet for a thirty hour period. The observations spanned the time leading up to superior conjunction and periastron, and continued for several hours thereafter:

The resulting time series looks like this:

The most remarkable feature of the light curve is the dip at time 2454424.72. The alignment of the planetary orbit turns out to be close enough to edge-on that a secondary eclipse occurs. The a-priori chance of observing the eclipse was only about 15%, and so we were lucky. Our interpretation of the light curve is that we’re seeing the planet heat up rapidly, from a temperature of roughly 800K to a temperature of about 1500K over a time period lasting roughly five or six hours. This indicates that the starlight is being absorbed at quite a high level in the atmosphere, where the air is thin and the heat capacity is low.

The details are all in the Nature paper. I’ll be posting it on astro-ph shortly, but in the meantime, a .pdf draft of the article is here, along with the (quite extensive) supplemental information section, and the figures (one and two) from the article.

The information that comes directly from Spitzer amounts to a 30-hour, one-pixel grayscale movie of a storm that was brewing on the planet back in the Monroe Administration. Hydrodynamical modeling, however, can flesh out the details, and the goal over the coming years will be to compute simulations that are as detailed and as physically correct as possible. In the next post, I’ll go into more detail, but here’s an advance look at the results of a “synthetic mission” in which a probe has been inserted into orbit around the planet 2.2 days prior to periastron. The resulting footage runs through 8.9 days after periastron. The orbital dynamics and the illumination are all self-consistent…

Footage from a synthetic probe.

Coming soon…

January 27th, 2009 1 comment

Download the poster-size version (4.4MB).


January 19th, 2009 6 comments

WASP-12b. Now there’s an unpleasant travel destination.

Nevertheless, this particular planet, whose transits were recently announced by the SuperWASP collaboration, is quite a remarkable world. For starters, inveterate bottle-poppers can celebrate a WASP-12b New Year on literally nine out of every ten days — the orbital period is a mere 26 hours and 11 minutes. The temperature of the planetary photosphere at the substellar point likely exceeds 2500K. Cherry orange, to be exact.

Because of its ultra-short orbital period, WASP-12b is attracting quite a bit of interest. The planet has a radius 1.8x larger than Jupiter, which should make it eminently feasible to detect secondary transits from the ground in either the optical or near-infrared. One expects, furthermore, that a planet with an orbital period just a shade over a day should have long since damped out its eccentricity, but (to better than 2-sigma) the orbit appears to be non-circular, with e=0.049 +/- 0.015. Even if another planet exists in the system, there should long since have been evolution to a tidal fixed point, followed by circularization. If the orbit really is eccentric, then GR precession of the periastron amounts to a whopping 0.2 degrees per year, nearly 2000x faster than Mercury’s stately 43” per century.

I got an opportunity to visit Harvard this month, and while I was there, David Latham remarked that he had used a remotely operated telescope in Arizona to get a high-precision light curve of a WASP-12b transit. Latham is a meticulous observer, and so, in order to get the best possible baseline, he had cued up the telescope a number of hours prior to the predicted ingress. He related that he’d been completely startled to find, upon analyzing his photometry, that the transit had occurred several hours ahead of schedule. Without a doubt, transit timing variations are going to be one of the big exoplanet stories of 2009, but they’re going to be measured in seconds, not hours. Imagine the commotion that would result if the Sun rises a few hours late tomorrow morning!

The WASP-12 mystery was solved by the amateur astronomers Veli-Pekka Hentunen and Markku Nissinen of Taurus Hill Observatory near Varkaus, Finland. Bruce Gary, who runs the Amateur Exoplanet Archive forwarded the news of their work:

AXA contributors and TransitingPlanets members,

I just received two data files for WASP-12 as observed by Veli-Pekka Hentunen and Markku Nissinen (Finland) which suggest that the discovery paper for this exoplanet has a misprint for the ephemeris. Their observations on January 1 was a “no show” (attached) whereas their observations on January 4 had a nice transit (attached). According to the discovery paper’s ephemeris there should have been a transit on January 1 but not on January 4. However, the discovery paper has a discrepancy between the stated ephemeris and the stated HJD for WASP survey observations. The Hentunen and Nissinen observations can be explained if the discovery paper’s stated WASP survey HJD is correct and their HJDo has a number transposition, such that HJDo = 4506.7961 (instead of 4506.9761). This is described on the AXA web page for WASP-12:


We amateurs have to keep the pro’s honest! Nice work, Veli-Pekka Hentunen and Markku Nissinen.

Bruce L. Gary, webmaster
Amateur Exoplanet Archive

Indeed! The typographical error in the discovery ephemeris has now been corrected, and with it, the puzzling “early” transit was revealed to be a completely separate event in the unending sequence of near-daily occultations. It seems somehow fitting that a seemingly alarming discrepancy for the hottest planet known was resolved by a pair of dedicated amateur observers during the long, dark, and frozen Finnish nights.

Categories: worlds Tags: , ,

HAT found a Neptune,

January 6th, 2009 2 comments

and at 880K it’s close to ten times hotter (but likely the same color) as the original edition.

In the twenty months following Gillon et al.’s startling discovery that Gliese 436b is observable in transit, literally dozens of additional transiting planets have been found. New transiting hot Jupiters are now routine enough that they’re generally trotted out in batches. Reported cases of transit fever have also been on the decline, with symptoms often amounting to little more than a passing distraction.

That said, it’s been been a very long dry spell waiting for a second example of a transiting Neptune-mass planet, which makes HAT-P-11b both exciting and newsworthy. In a preprint that muscled its way to the top of today’s astro-ph mailing, Gaspar Bakos and collaborators have produced a admirably solid analysis of what’s definitely the toughest ground-based detection to date.

HAT-P-11b’s transit depth is 4.2 millimag, which is the smallest planet-produced dip yet detected by a photometric survey. (HD 149026b has a smaller transit depth, but it was discovered via the Doppler velocity method and then followed up photometrically for the transits during the time windows predicted by the orbital solution.) The HAT-P-11b analysis was further confounded by a photometrically variable parent star and ~5m/s stellar jitter on the radial velocity observations. The paper is definitely worth reading carefully.

HAT-P-11b is quite similar in mass and radius to Gliese 436b, and it’s actually somewhat larger than Neptune on both counts. When the mass and radius are compared to theoretical models, it’s clear that, like Gliese 436, it’s mostly made of heavy elements (that is, some combination of metal, rock and “ice”) with an envelope of roughly 3 Earth masses of hydrogen and helium). It’s completely dwarfed when placed next to an inflated hot Jupiter, HAT-P-9b, for instance:

Interestingly, HAT-P-11b seems to have a significant eccentricity, on the order of e=0.2. Drawn to scale with the parent star, the orbit looks like this:

The dots demarcating the orbit are not to scale. With 500 pixels of resolution, you can just barely see the planet. (I put one in front of the star, and tacked a copy onto the orbit for good measure.)

The e=0.15 eccentricity of Gliese 436b has caused a lot of consternation. For any reasonable value of the so-called tidal quality factor, Q, the circularization timescale for Gliese 436b’s orbit is considerably shorter than the age of the system. This has led to attempts (to date unfulfilled) to locate Gliese 436c. HAT-P-11b doesn’t have this problem. For a given Q, it’s circularization timescale is a full thirty times longer than that of 436b. The orbit will still be measurably eccentric even when the 0.8 solar mass primary starts to turn into a red giant.

Categories: worlds Tags: , , ,