extrasolar planets - queen's universityastro.queensu.ca/~tjb/het603/extrasolar.pdf · observe...
TRANSCRIPT
ExtraSolar Planets
ExtraSolar Planets● We have estimated there may be 1020 billion stars in Milky Way with Earthlike planets, hospitable for life. But what evidence do we have that such planets even exist?● 11 years ago we knew of no planets outside our own Solar System (aside from 1 or 2 planets around pulsars, not interesting in terms of life). We now know of 185 extrasolar planets (Mar 2006)
Planets and Brown Dwarfs
● We need to have a clear definition of what a planet is. We need to consider lowmass stars and brown dwarfs (failed stars)
(1) Stars: we define a star as an object massive enough to burn H in its core. This requires a mass > 0.08 solar masses
(2) Brown Dwarfs: These are objects which formed similar to stars, but not big enough to fuse H. They can burn deuterium (D).
● Not clear how small brown dwarfs could be; could they be as small as “planets”? How do we distinguish planets and BDs? They form differently but this is hard to tell
Decision: BDs are those objects big enough to burn D. BDs thus have masses > 13 M
jup and planets have masses < 13 M
jup
Ways to Detect ExtraSolar Planets
● There are several possible ways to find ESPs:
● Detection of radio messages from intelligent civilizations: we will discuss this later (nothing so far!)
● Observe the planets directly (i.e. take a picture)
● Planetary transits: planet blocks off some of star's light
● Gravitational microlensing
● Wobbles in star's position or velocity caused by planet
ExtraSolar Planets: The Doppler Technique
● Nearly all ESPs discovered using the Doppler technique (173)
● A star and giant planet both orbit together around their common center of mass. Star is much more massive, so star's orbit much smaller than planet's.
Orbital “Wobbles”
● So a giant planet will cause the star's orbit to wobble. The more massive the planet, the larger the wobbles.
● We will talk later about astrometric detections of such wobbles, which have not yielded results to date.
Doppler Technique
● Look for “wobbles” in radial velocity of the star, by Doppler shifting of its light (compare spectroscopic binary stars).
Doppler Technique
● Doppler technique only measures that part of star's velocity along the line of sight
● The changes in velocity are small and hard to measure, between few m/sec to ~100 m/sec
● Requires special spectroscopic setup: high spectroscopic resolution to measure such small velocities
● 1995: Swiss team (Michel Mayor & Didier Queloz) announced the first extrasolar planet (51 Pegasus). An American team (Marcy, Butler & co) have found most extrasolar planets so far
What Doppler Technique Tells Us
(1) Period of the Doppler shifting gives us the orbital period (of star and planet)
(2) Shape of the lightcurve gives us the orbital eccentricity
Nearly Circular Orbit
Very Elliptical/Eccentric Orbit
(3) Period + Star Velocity + Star Type gives: Planetary orbital radius Planet velocity Planet's Mass (nearly!): Msini, where i = inclination
(4) Combined with transit data also get: definitive planet mass, plus planet size and density
Gravitational Lensing
● Gravity bends light. Spectacular gravitational lensing seen in galaxy clusters
Gravitational Microlensing
● Less spectacular microlensing can be seen when objects such as brown dwarfs or planets pass in front of background stars
● Duration depends on planet's mass and speed. So by measuring duration we find out the planet's mass
Durations of a few hours for Jupitersized planets● Each event is onetime only: you don't get a second chance! No information about planet's orbit (just its size)● But microlensing is sensitive to Earthmass objects (unlike Doppler technique), with distances from 15 AU
● A variation of this method is to search for secondary spikes in microlensed stars, caused by planets orbiting them
● lots of searches underway around the world, e.g. PLANET
● PLANET found that no more than 1/3 of all sunlike stars have Jupitersized planets with orbits between 1.5 – 4 AU (2001)
First Success!!!
Probably a planet ~2 MJup
a distance of ~3 AU from its star
More Microlensing Discoveries ...
OGLE2005390Lb: Mp = 310 M
E
a=24 AU D=6.6 kpc Ms = 0.22 M
sun
Tp ~ 50K (like Pluto!)
OGLE2005BLG169b: Mp = 13 M
E
a=3 AU Tp ~ 70K (cold!)
● 4 microlensed planets now known● Two of these (below) are several Earth masses, orbits of a few AU such planets should be common, and detectable via microlensing● Planets common around fainter M stars?
Planetary Transits:
● A planet passing in front of a star can block and dim some of the star's light, something like 12% for a “hot Jupiter”, with durations of typically hours
● This is similar to studies of eclipsing binary stars: can get planet's size, distance from star, and orbital period. With velocity measurements, could then get planet's mass and hence density (rocky, gas giant?) Transit Animation
Advantages of Transits:
● Sensitive to Earthsized planets, unlike most other methods. Better than microlensing, because you can followup
● The geometry is known (edgeon), simplifying things
Disadvantages of Transits:
● Planet orbit has to be edgeon to us to see transit. This will be rare, so lots of stars have to be monitored
● The brightness dip is small, so difficult to measure
Transit Results to Date:
● First transit seen (1991) was that in HD209458, a planet found earlier using the Doppler technique.
● With the velocity data, the planet's radius, mass, and density could be determined: it is definitely a gas giant
● HST spectra found Na, H, O, and C in the planet's upper atmosphere, which is escaping from the star (because the planet is so near its star, and thus so hot)
HD209458
Further Transit Discoveries● 5 found by Optical Gravitational Lensing Experiment (OGLE)
● These have masses 0.51.5 MJ, sizes ~ Jupiter, a= 0.020.05 AU, periods 14 days, T: 10002000 deg “Hot Jupiters”
● Many transit experiments in progress, and planned space missions (later)
TrES1● Found in 2003 by STARE project using 10cm telescope!
● M = 0.75 MJ, a = 0.04 AU, P = 3 days (Hot Jupiter)
● Spitzer Space Telescope detected IR photons from this planet
Transit Searches in Globular Clusters
● 47 Tuc has been searched for planets by HST (Gilliland et al. 2000). None found! ● But they only searched a small region near the cluster center● I'm involved in a search for planets over a much larger area in two GCs: in progress but no planets so far! ● Looks like stellar metallicity is very important for planet formation
Proper Motion (astrometry):● Stars that are close enough to us to have observable proper motions are candidates. Idea is to look for wobbles in their motions across the sky caused by a massive planet orbiting them
Proper Motion
● Very difficult to measure because wobbles are small. Most sensitive to massive planets
● But planned for groundbased interferometers. E.g. Keck hopes to detect wobbles < 3000 km at the distances of the nearest stars. Could detect Uranussize planets around stars up to 60 light years away. Also VLT
● But no planets discovered yet ...
● Space Interferometry Mission (SIM) should be able to detect Earthsize planets! Also GAIA.
Imaging ExtraSolar Planets
● Observing planets directly is hard! Planets shine mostly by reflected light
Planets are ~1 billion times fainter than star Planets are very close to their stars (1 AU at 1 pc is 1” in angular size (atmospheric resolution limit).
● Need very high angular resolution and blocking of light from star. Possible from the ground with extreme adaptive optics and a coronagraph (e.g. Gemini), or interferometery. Even better in space (e.g. JWST, Terrestrial Planet Finder, SIM)
● Works best in the IR, where the contrast between star and planet is lowest, and with smaller, fainter stars
Success: 2M1207b● Imaged with the VLT in 2004 and confirmed in 2005. Star is (faint) Brown Dwarf● 5± 1 M
J ;1.5 R
J , 41 AU, D=53 pc
● Spectrum shows evidence for water absorption
Two Other Possibilities
AB Pictoris13.5 ± 0.5 M
J, a = 275 AU
GQ Lupi22 ±21 M
J a = 103 AU R = 2R
J
A “niche” for each technique
C. Lineweaver
Properties of Extrasolar Planets
● As of Mar 2006, 185 planets have been found around ~150 stars, 149 of these planets using the Doppler technique
At least 10% of stars surveyed have detected planets (fraction depends on stellar metallicity see below)
Orbital periods from few to thousands of days!
●18 stars have multiple planets (2 or more planets)
● Almost all giant planets: most techniques are sensitive to massive planets close to their stars (Earthmass planets difficult at present time). This is an important selection effect we have to bear in mind.
Host Star Properties● Stars richer in heavy elements are more likely to host planets● Most planets have been found around Sunlike stars (F, G, K) but this is partly a selection bias: planets are now found around fainter M stars● 18 systems are multiplestar (2 or more stars): it is possible to have planets in such systems (but are these planets habitable?)
ESP Masses
ESP Sizes and Densities● Density: 0.1 to 2.5 times that of Jupiter: Not silicateiron composition● But not many ESPs have measured radii (transits, microlensing)
ESP SemiMajor Axes
ESP SemiMajor Axes
● Most ESPs are very close to their host stars!● Compare Mercury at 0.39 AU● “Hot Jupiters”!
● Many ESPs are on very eccentric orbits!● Unlike our solar system● Not good for life: extreme hot/cold cycles
ESP Orbital Eccentricities
Puzzles
● Our planet formation model predicts nearcircular orbits: planets form from condensations in rotating disk of gas and dust
● Compare our solar system: gas giants much further out partly a selection effect: most sensitive to massive, inner
planets; but will improve with time
● Still, this doesn't explain everything: how did these planets get so close? Unlikely they could have formed so close.
just too hot for material to condense to form gas giants
● Current thinking: They formed out at several AU, then migrated inward due to tidal/friction effects in solar nebula
● Type I migration: interaction between giant planet and circumstellar gas/dust disk pushes planet inwards
● Type II migration: Gap in disk opens and migration slows
● Have to halt the process: removal of disk; tidal/magnetic interactions between planet/star/disk
● Multiplegiant cases can explain higheccentricity orbits by resonances or close encounters between giants
● Not clear how difficult it is for Earthmass planets to form and survive under giant planet migration
● Quite likely every planetary system has lost planets ...
Type II Migration
Implications For finding Earthlike Planets
● Having Jupitersized planets in elliptical orbits, near their host stars really decreases chances of forming Earthsized planets
● We need: Jupitersized planets in Jupiter orbits (circular and >5 AU from their stars)! Only then can we be (more) secure that Earthsized planets can form within the habitable zone
● Doppler technique biased against finding such planets: ones closer to their stars having faster periods and easier to detect
● However, ~12+ years on, we are now starting to find such Jupiter analogues!
55 Cancri: a 3planet system!
(1) 15 days, 0.84 MJ, 0.115 AU
(2) 44 days, 0.21 MJ, 0.24 AU
(3) 14 yrs!, 4 MJ, 5.9 AU
● Compare Jupiter at 5.2 AU, e = 0.04, P = 11.9 years
55 Cancri(AAT Planet Search Program)
Some Thoughts
● Planets could still exist outside giant planets (but far from star)
● Earthlike planets may be captured by giant planets (like Trojan asteroids), and go along for the ride to inner solar system
● Moons of the giant planets may be suitable for life
● We have always to be aware of the selection techniques in the Doppler technique: we are discovering more planets as time goes on. At least ~10% of stars studied so far have planets, and this fraction will only increase with time
● So it's impossible to estimate number of Earthlike planets in our Galaxy from the Doppler data
Ways to Find EarthMass Planets● Doppler technique (on current telescopes) will never find Earthlike planets: their velocity wobbles are simply too low
We need different techniques: we have discussed two already that should yield Earthsize planets: transits and microlensing
lots of groundbased programs at the moment
● Spacebased missions hold much promise:
Transit: COROT, Eddington, Kepler (within next few years)
Interferometry: SIM, TPF, Darwin (within next 1520 years)
● The interferometry instruments and/or large groundbased telescopes should be able to directly detect extrasolar planets and life gases in their atmospheres!
Kepler (2007)
SIM (2009)
Terrestrial Planet Finder (TPF) and
Darwin (Next Decade)
Detecting Life on Exoplanets?● Direct sampling: send a probe! Pretty damn hard ...
● IR spectrum of exoplanet gives temperature of surface and/or atmosphere, and atmospheric composition
If H2O vapour and CO
2 found, and if temp right for liquid
water and carbon compounds conditions good for LAWKI
● Strong O3 (ozone) would indicate O
2 and a biosphere (e.g.
oxygenic photosynthesis).
● Even stronger evidence: existence of redox pairs such as O2
and CH4 (i.e. out of chemical equilibrium)
lack of O2 does not mean there's no biosphere!
● Spectra might also detect atmospheric gases or effects of chlorophyll, or by changes in the lightcurve with time