Transcript No Slide Title

DEPARTMENT OF PHYSICS AND ASTRONOMY
Life in the Universe:
Extra-solar planets
Dr. Matt Burleigh
www.star.le.ac.uk/~mbu
3677 Timetable
• Today and Tuesday 11am: MB Extrasolar
planets
• Then Mark Sims (Life in the solar system)
Dr. Matt Burleigh
3677: Life in the Universe
Contents
• Methods for detection
– Doppler “wobble”
– Transits
– Direct Imaging
• Characterisation
– Statistics
– Implications for formation scenarios
Dr. Matt Burleigh
3677: Life in the Universe
Useful reading / web sites
• 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
Dr. Matt Burleigh
3677: Life in the Universe
What is a planet?
• 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!
Dr. Matt Burleigh
3677: Life in the Universe
A brief history of exoplanets
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Radial Velocity Technique
(Doppler “Wobble”)
• Star + planet orbit
common centre of
gravity
• As star moves
towards observer,
wavelength of light
shortens (is blueshifted)
• Light red-shifted
as star moves
away
Dr. Matt Burleigh
3677: Life in the Universe
Measuring Stellar Doppler shifts
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Measuring Stellar Doppler shifts
• 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
Dr. Matt Burleigh
3677: Life in the Universe
• N.B. M* comes from
the spectral type
Dr. Matt Burleigh
3677: Life in the Universe
Doppler Wobble Method: Summary
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Doppler Wobble Method: Summary
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Transits
• Planets observed at inclinations near 90o will transit their
host stars
Dr. Matt Burleigh
3677: Life in the Universe
Transits
• 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)
Dr. Matt Burleigh
3677: Life in the Universe
Transits
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
Dr. Matt Burleigh
3677: Life in the Universe
Transits
• 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!
Dr. Matt Burleigh
3677: Life in the Universe
Transits
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Transits
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Super WASP
• Wide Angle Search for Planets
(by transit method)
• First telescope located in La
Palma, second in South Africa
• Operations started May ‘04
• Data stored and processed at
Leicester
• >20 new planets detected!
• www.superwasp.org
• www.wasp.le.ac.uk
Dr. Matt Burleigh
3677: Life in the Universe
Direct detection
• Imaging = spectroscopy = physics:
composition & structure
• Difficult
• 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”
Dr. Matt Burleigh
3677: Life in the Universe
Direct detection
• 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)
Dr. Matt Burleigh
3677: Life in the Universe
First directly imaged planet?
• 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!
Dr. Matt Burleigh
3677: Life in the Universe
First directly imaged planets
• 3 planets around
HR8799, 130 light
years away (40pc)
• Young (60million years
old)
• Three planets at 24, 38
and 68AU separation
– In comparison, Jupiter is
at 5AU and Neptune at
30AU)
• Masses of 7Mjup,
10Mjup and 10Mjup
• Also, a 3Mjup planet
around Fomalhaut, at a
separation of 120AU
Dr. Matt Burleigh
3677: Life in the Universe
Direct detection:
White Dwarfs
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Direct Detection: White Dwarfs
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
Dr. Matt Burleigh
3677: Life in the Universe
What we know about
extra-solar planets
• 328 planets now found
• 34 multiple systems
• 52 transiting planets – can
directly measure radii
• Unexpected population with
periods of 2-4 days: “hot
Jupiters”
• Planet with orbits like Jupiter
discovered (eg 55 Cancri d)
• Is our solar system typical?
Dr. Matt Burleigh
3677: Life in the Universe
Extra-solar planet period distribution
• Notice the “pile-up”
at periods of 2-4
days / 0.04-0.05AU
• The most distant
planets discovered
by radial velocities
so far are at 5-6AU
• Imaging surveys
finding very wide
orbit planets
Dr. Matt Burleigh
3677: Life in the Universe
Extra-solar planet mass distribution
• Mass distribution peaks at 12 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?
Dr. Matt Burleigh
3677: Life in the Universe
Selection effects
• 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, as is the Transit method
– Imaging surveys sensitive to massive planets in very wide orbits
(>10AU)
• These methods are not yet sensitive to planets as small
as Earth, even close-in
• As orbital period increases, the Doppler Wobble method
becomes insensitive to planets less massive than
Jupiter
• The length of time that the DW 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
– But imaging surveys can find the widest planets
Dr. Matt Burleigh
3677: Life in the Universe
What we know about extra-solar planets:
Mass versus semi-major axis
• Blue – exoplanets
• Red – solar system
• Many of the known solar
systems have ~Jupitermass planets in small
orbits, <0.1AU
– Selection effect of Doppler
surveys
• But almost no superJupiters are found in close
orbits
– Real, not a selection effect
Dr. Matt Burleigh
3677: Life in the Universe
What we know about extra-solar planets
Eccentricity vs semi-major axis
:
extra-solar planets
solar system planets
Dr. Matt Burleigh
observational bias
- large distribution of e
(same as close binaries)
- 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
- planets on circular orbits do exist far
away from star
- the planets in our own system have
small eccentricities ie STABLE
3677: Life in the Universe
Statistics of the Doppler Wobble surveys:
Summary
• 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
– 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
Dr. Matt Burleigh
3677: Life in the Universe
What about the stars themselves?
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Metallicity
The abundance of elements
heavier than He relative to the Sun
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Metallicity
• 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 millisecond 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
Dr. Matt Burleigh
3677: Life in the Universe
Planet formation
scenarios
• 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.
Dr. Matt Burleigh
3677: Life in the Universe
Gas accretion onto cores
• 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 nongravitational 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.
Dr. Matt Burleigh
3677: Life in the Universe
Gravitational instability
• 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.
Dr. Matt Burleigh
3677: Life in the Universe
Where do the hot Jupiters come from?
• 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 protoplanetary disk, but generally >1AU
• Hot Jupiters currently at ~0.03-0.04AU cannot
have formed there
• Migration!
Dr. Matt Burleigh
3677: Life in the Universe
Planetary 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
Dr. Matt Burleigh
3677: Life in the Universe
Planetary migration & terrestrial planets
• 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
Dr. Matt Burleigh
3677: Life in the Universe
The future: towards other Earths
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Towards other Earths
Telescope
Method
Date
Corot
Transits
2007
Kepler
Transits
2008
GAIA
Astrometry
2012
SIM
Interferometry 2012-15 (?)
Plato
Transits
Darwin/TPF
Interferometry 2020+ (?)
50m ELT
Imaging
Dr. Matt Burleigh
2017
2019
3677: Life in the Universe
Towards Other Earths: Habitable Zones
Left: courtesy Prof. Keith Horne, St.Andrews
Right: courtesy Prof. Barry Jones, Open
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Towards Other Earths: Biomarkers
• 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
Dr. Matt Burleigh
3677: Life in the Universe
Dr. Matt Burleigh
3677: Life in the Universe