Life in the Universe: Extra-solar planets - PowerPoint PPT Presentation

Life in the Universe:Extra-solar planets

Dr. Matt Burleigh

www.star.le.ac.uk/~mbu

• Today, and next Friday: MB Extrasolar planets
• Then Derek Raine
• After Xmas: Mark Sims (Life in the solar system)

• Methods for detection
• Doppler “wobble”
• Transits
• Microlensing
• Direct Imaging
• Characterisation
• Statistics
• Implications for formation scenarios

• Nature, Vol. 419, p. 355 (26 September 2002)
• Extra-solar planets encyclopaedia
• California & Carnegie Planets Search
• How stuff works planet-hunting page
• Includes lots of animations & graphics
• JPL planet finding page
• Look at the science & multimedia gallery pages

• International Astronomical Union definition –
• An object orbiting a star
• But see later this lecture…
• Too small for dueterium fusion to occur
• Less than 13 times the mass of Jupiter
• Formation mechanism?
• Forms from a circumstellar disk
• Lower mass limit – IAU decided last year that Pluto should be downgraded!

• 1991 Wolszczan & Frail discovered planets around a pulsar PSR1257+12
• Variations in arrival times of pulses suggests presence of three or more planets
• Planets probably formed from debris left after supernova explosion
• 1995 Planets found around nearby Sun-like star 51 Peg by Doppler “wobble” method
• Most successful detection method by far
• 265 exoplanets found to date

• Star + planet orbit common centre of gravity
• As star moves towards observer, wavelength of light shortens (is blue-shifted)
• Light red-shifted as star moves away

• Method:
• Observe star’s spectrum through a cell of iodine gas
• Iodine superimposes many lines on star’s spectrum
• Measure wavelength (or velocity) of star’s lines relative to the iodine

• Method:
• Measure wavelength (or velocity) of star’s lines relative to the iodine
• Dl / le = (l0-le) / le = vr / c

lo=observed wavelength, le=emitted wavelength

• Precision of current surveys is now 1m/s:
• Jupiter causes Sun’s velocity to vary by 12.5m/s
• All nearby, bright Sun-like stars are good targets
• Lots of lines in spectra, relatively inactive
• Limited to gas planets and larger
• Note recently discovered “hot Neptunes” (>14MEarth)
• Not yet suitable for Earth-like planets
• Length of surveys limits distances planets have been found from stars
• Earliest surveys started 1989
• Jupiter (5AU from Sun) takes 12 yrs to orbit Sun
• Saturn takes 30 years
• Would remain undetected
• Do not see planet directly

• Since measure K (= v* sin i), not v* directly, only know mass in terms of the orbital inclination i
• Therefore only know the planet’s minimum mass
• If i=90o (eclipsing or transiting) then know mass exactly

Orbital plane

i=900

Orbital plane

i0

• Planets observed at inclinations near 90o will transit their host stars

• Assuming
• The whole planet passes in front of the star
• And ignoring limb darkening as negligible
• Then the depth of the eclipse is simply the ratio of the planetary and stellar disk areas:
• i.e. Df / f* = pRp2 / pR*2 = (Rp / R*)2
• We measure the change in magnitude Dm, and obtain the stellar radius from the spectral type
• Hence by converting to flux we can measure the planetary radius
• Rem. Dm = mtransit – m* = 2.5 log (f* / ftransit)
• (smaller number means brighter)

Example: first known transiting planet HD209458b

• Dm = 0.017 mags
• So (f* / ftransit) = 1.0158, i.e. Df=1.58%
• From the spectral type (G0) R=1.15Rsun
• So using Df / f* = (Rp / R*)2 and setting f*=100%
• Find Rp=0.145Rsun
• Since Rsun=9.73RJ then
• Rp = 1.41RJ

• HD209458b more:
• From Doppler wobble method know M sin i = 0.62MJ
• Transiting, hence assume i=90o, so M=0.62MJ
• Density = 0.29 g/cm3
• c.f. Saturn 0.69 g/cm3
• HD209458b is a gas giant!

• For an edge-on orbit, transit duration is given by:
• Dt = (PR*) / (pa)
• Where P=period in days, a=semi-major axis of orbit
• Probability of transit (for random orbit)
• Ptransit= R* / a
• For Earth (P=1yr, a=1AU), Ptransit=0.5%
• But for close, “hot” Jupiters, Ptransit=10%
• Of course, relative probability of detecting Earths is lower since would have to observe for up to 1 year

• Advantages
• Easy. Can be done with small, cheap telescopes
• E.g. WASP,
• Possible to detect low mass planets, including “Earths”, especially from space (Kepler mission, 2008)
• Disadvantages
• Probability of seeing a transit is low
• Need to observe many stars simultaneously
• Easy to confuse with starspots, binary/triple systems
• Needs radial velocity measurements for confirmation, masses

• Wide Angle Search for Planets (by transit method)
• First telescope located in La Palma
• Operations started May ‘04
• Data stored at Leicester
• Three new planets detected!
• www.superwasp.org
• www.wasp.le.ac.uk

• A consequence of general relativity
• The grav. field of a relatively nearby star can bend the light of a more distant object as it passes in front of it, as seen from Earth
• The star doing the lensing brightens as a result
• We record this brightening, which can last for days
• If the lensed star has a planetary companion, the characteristic lensing light curve is modified
• Signals from an Earth-like planet would be strong (>5%) but brief (few hours)
• 4 planets found so far, including one at 5.5 Earth masses!

• Imaging = spectroscopy = physics: composition & structure
• No planet in orbit around another star has been directly imaged
• Why?
• Stars like the Sun are billions of times brighter than planets
• Planets and stars lie very close together on the sky
• At 10pc Jupiter and the Sun are separated by 0.5”

• Problem 1:
• Stars bright, planets faint
• Solution:
• Block starlight with a coronagraph
• Problem 2:
• Earth’s atmosphere distorts starlight, reduces resolution
• Solution:
• Adaptive optics, Interferometry – difficult, expensive
• Or look around very young and/or intrinsically faint stars (not Sun-like)

• 2M1207 in TW Hya association
• Discovered at ESO VLT in Chile
• 25Mjup Brown dwarf + 5Mjup “planet”
• Distance ~55pc
• Very young cluster ~10M years
• Physical separation ~55AU
• A brown dwarf is a failed star
• Can this really be called a planet?
• Formation mechanism may be crucial!

• White dwarfs are the end state of stars like the Sun
• 1,000-10,000 times fainter than Sun-like stars
• contrast problem reduced
• Outer planets should survive evolution of Sun to white dwarf stage, and migrate outwards
• more easily resolved
• Over 100 WD within 20pc
• At 10pc a separation of 100AU = 10” on sky
• I have a programme to search for planets around nearby WD with the Gemini 8m telescopes
• We call it “DODO” – Degenerate Objects around Degenerate Objects or Dead Objects etc

Proper motions

Two images taken one year apart

• The faint objects in this field could be massive planets in wide orbits around this nearby white dwarf
• The white dwarf moves relatively quickly compared to background stars in the field (see movie)
• If a faint object moves with the WD, then I would get excited
• But in this case, there is nothing, but we could have detected something as small as ~5MJup!
• www.le.ac.uk/~mbu

• 265 planets now found
• 25 multiple systems
• 33 transiting planets – can directly measure radii
• Unexpected population with periods of 3-4 days: “hot Jupiters”
• New population from transit surveys at ~2 days
• First planet with an orbit like Jupiter discovered (55 Cancri d)
• Is our solar system typical?

• Notice the “pile-up” at periods of 3-4 days / 0.04-0.05AU
• The most distant planets discovered so far are at 5-6AU
• New discovery of transiting planets at ~2 days

• Mass distribution peaks at 1-2 x mass of Jupiter
• Lowest mass planet so far: 5.5xMEarth
• Super-Jupiters (>few MJup) are not common
• Implications for planet formation theories?
• Or only exist in number at large separation?
• Or exist around massive stars?

• Astronomical surveys tend to have built in biases
• These “selection effects” must be understood before we can interpret results
• The Doppler Wobble method is most sensitive to massive, close-in planets
• It is not yet sensitive to planets as small as Earth, even close-in
• As orbital period increases, the method becomes insensitive to planets less massive than Jupiter
• The length of time that the surveys have been active (since 1989) sets the upper orbital period limit
• Only now are analogues of Jupiter in our own Solar System going to be found

• Blue – exoplanets
• Red – solar system
• Many of the known solar systems have ~Jupiter-mass planets in small orbits, <0.1AU
• Selection effect of Doppler surveys
• But almost no super-Jupiters are found in close orbits
• Real, not a selection effect

:

- large distribution of e

(same as close binaries)

What we know about extra-solar planets

Eccentricity vs semi-major axis

- most extra-solar planets are on orbits

much more eccentric than the giant

planets in the solar system: bad news for survivability of terrestrial planets

- planets close to the star are tidally circularized

observational bias

extra-solar planets

- planets on circular orbits do exist far

away from star

- the planets in our own system have

small eccentricities ie STABLE

solar system planets

• Of 2000 stars surveyed
• ~5% have gas giants between 0.02AU and 5AU
• Trends suggest ~10% of stars have planets in orbits 5-7AU
• 0.85% have hot Jupiters
• Real effect
• Hot Jupiters are not massive
• Almost all have Msini~1Mjup or less
• No close-in, “super-Jupiters”
• Mass distribution strongly peaks at 1Mjup and falls as dN/dM~M-0.7
• But surveys currently biased towards hot Jupiters
• Expect mass distribution to flatten somewhat as long periods, super-Jupiters are discovered

• Surveys began by targeting sun-like stars (spectral types F, G and K)
• Now extended to M dwarfs
• Incidence of planets is greatest for late F stars
• F7-9V > GV > KV > MV
• Few low mass M dwarfs known to have a planets despite ease of detectability
• Stars that host planets appear to be on average more metal-rich

• Overall, ~5% of solar-like stars have radial velocity –detected Jupiters
• But if we take metallicity into account:
• >20% of stars with 3x the metal content of the Sun have planets
• ~3% of stars with 1/3rd of the Sun’s metallicity have planets

• Does this result imply that planets more easily form in metal-rich environments?
• If so, then maybe planet hunters should be targeting metal-rich stars
• Especially if we are looking for rocky planets
• This result also implies that chances of very old lifeforms (> few billion years) in the Universe are slim
• With less heavy elements available terrestrial planets may be smaller and lower in mass than in our solar system
• Is there a threshold metallicity for life to start (e.g. ½ solar)?
• BUT Sigurdsson et al. (2003, Science, 301, 193) claim that a milli-second pulsar in globular M4 has a Jupiter size companion
• Claim based on timing anomalies
• If true, then planets may have been forming 12 billion years ago in a very metal-poor environment (<0.1 x solar)
• Alternatively, planet may have formed from debris of supernova explosion that created the pulsar
• Or planet does not exist, timing anomalies have another cause

• There are two main models which have been proposed to
• describe the formation of the extra-solar planets:
• Planets form from dust which agglomerates into cores which then accrete gas from a disc.
• A gravitational instability in a protostellar disc creates a number of giant planets.
• Both models have trouble reproducing both the observed distribution of extra-solar planets and the solar-system.

• Planetary cores form through the agglomeration of dust into grains, pebbles, rocks and planetesimals within a gaseous disc
• At the smallest scale (<1 cm) cohesion occurs by non-gravitational forces e.g. chemical processes.
• On the largest scale (>1 km) gravitational attraction will dominate.
• On intermediate scales the process is poorly understood.
• These planetesimals coalesce to form planetary cores and for the most massive cores these accrete gas to form the giant planets.
• Planet formation occurs over 107 yrs.

• A gravitational instability requires a sudden change in disc properties on a timescale less than the dynamical timescale of the disc.
• Planet formation occurs on a timescale of 1000 yrs.
• A number of planets in eccentric orbits may be formed.
• Sudden change in disc properties could be achieved by cooling or by a dynamical interaction.
• Simulations show a large number of planets form from a single disc.
• Only produces gaseous planets – rocky (terrestrial) planets are not formed.
• Is not applicable to the solar system.

• No element will condense within ~0.1AU of a star since T>1000K
• Planets most likely form beyond the “ice-line”, the distance at which ice forms
• More solids available for building planets
• Distance depends on mass and conditions of proto-planetary disk, but generally >1AU
• Hot Jupiters currently at ~0.03-0.04AU cannot have formed there
• Migration!

• Planets migrate inwards and stop when disk is finally cleared
• If migration time < disk lifetime
• Planets fall into star
• Excess of planets at 0.03-0.04AU is evidence of a stopping mechanism in some cases
• Nature of stopping mechanism unclear: tides? magnetic cavities? mass transfer?
• Large planets will migrate more slowly
• Explanation for lack of super-Jupiters in close orbits

• Migrating giant planets will be detrimental to terrestrial planet survivability, if they both form at same time
• Planets interior to a migrating giant planet will be disrupted and lost
• Of course, these small planets may also migrate into star!
• If terrestrial planets can only survive when migration doesn’t take place through their formation zone (few AU),
• then 3%-20% of planet forming systems will possess them
• Alternatively, terrestrial planet formation may occur after dissipation of gas in proto-planetary disk (after 107 years)
• Disruption by a migrating giant planet unlikely
• Almost all planet-forming stars will have terrestrial planets

• Pace of planet discoveries will increase in next few years
• Radial velocity surveys will reveal outer giant planets with long periods like our own Solar System
• Transit surveys will reveal planets smaller than Saturn in close orbits
• First direct images will be obtained
• But the greatest goal is the detection of other Earths

• Habitable zone defined as where liquid water exists
• Changes in extent and distance from star according to star’s spectral type (ie temperature)
• It is possible for rocky planets to exist in stable orbits of habitable zones of known hot Jupiter systems
• If they were not previously cleared out by migration

Left: courtesy Prof. Keith Horne, St.Andrews

Right: courtesy Prof. Barry Jones, Open

• So we find a planet with the same mass as Earth, and in the habitable zone:
• How can we tell it harbours life?
• Search for biomarkers
• Water
• Ozone
• Albedo

• Dust disk around Fomalhaut
• Sub-mm image taken from James Clarke Maxwell telescope on Hawaii
• Disk has a hole in centre like a doughnut
• Part of disk appears to be perturbed

• Disk is 200 million years old
• Like the early Solar System
• Is a planet perturbing the disk & forming the hole?

• Young, Sun-like star only 3pc away
• Dust disk is clumpy
• Clumps seen to rotate
• Requires presence of at least one Jovian planet
• 1.55MJup companion confirmed by HST measurements