Host stars of exoplanets
Matthias Ammler-von Eiff
(TLS Tautenburg)
Literature - a selection
• Books:
– Perryman, M.: “The Exoplanet Handbook”, Cambridge, Cambridge
University Press, 2011 (topic of lecture: Chap. 8)
– Gray, D.F.: „The observation and analysis of stellar photospheres“,
3rd edition, Cambridge: Cambridge University Press, 2005
(advanced)
– Gray, R.O.: „Stellar spectral classification“, Princeton and Oxford:
Princeton University Press, 2009 (advanced)
– Schrijver, C.J. & Zwaan, C.: „Solar and stellar magnetic activity“,
Cambrigde: Cambridge University Press, 2000 (advanced)
– Stix, M.: „The Sun - An Introduction“, 2. korr. Druck, Berlin,
Heidelberg: Springer-Verlag, 1991 (advanced)
– Strassmeier, K.: „Aktive Sterne - Laboratorien der Astrophysik“,
Wien: Springer, 1997 (basic)
– Kaler, J.B.: „Sterne und ihre Spektren“, Heidelberg: Spektrum,1994
– Unsöld, A. & Baschek, B.: „Der neue Kosmos“, 6. Auflage, Berlin:
Springer, 1999 (basic)
Literature - a selection
• Research articles related to the subject:
– access arXiv e-prints via http://cdsads.ustrasbg.fr/abstract_service.html
– CoRoT-7: Bruntt, H. et al. (2010), A&A 519, 51, arXiv:1005.3208
– CoRoT-19: Guenther, E.W. et al. (2011), arXiv:1112.1035
– Kepler-22: Borucki, W.J. et al. (2012), ApJ 745, 120B,
arXiv:1112.1640
Aspects - an incomplete view
Planet detection
Detection techniques are indirect. It is the
star that is measured!
Planet properties
The accuracy of measured parameters depends
on the knowledge of the star!
Planet formation
Planet frequency and properties seem to depend on star!
Star-planet interaction
Can a planet cause features on the star?
Habitability
Existence of life depends on the star! And on the
location in the Galaxy? Other galaxies?
Contents of this lecture
• Determination of stellar parameters
– fundamental stellar parameters and their
determination
• Effects of metallicity
– correlation of stellar metallicity and planet occurence
• Stellar activity
– magnetic and chromospheric activity
– measurement of chromospheric activity
– stellar flares and their influence on habitable planets
I. DETERMINATION OF STELLAR
PARAMETERS
Determination of stellar parameters
• Most planets are found indirectly by studying the
light of the central star!
• We need to know the star in order to know its
planets!
• Methods:
–
–
–
–
–
photometry
spectroscopy
asteroseismology
interferometry
astrometry
Example: CoRoT-19b
Example: CoRoT-19b
Example: CoRoT-19b
Spectroscopic analysis
Example: CoRoT-19b
light curve
Example: CoRoT-19b
evolutionary models
Stellar Parameter Network
(incomplete!)
chemical
abundances
Spectroscopy
age
brightness
Asteroseismology
surface
gravity
rotational
velocity
luminosity
effective
temperature
radius
distance
Astrometry
rotational
period
model
isochrones
angular
diameter
Photometry
Interferometry
mass
mean stellar
density
Stellar Parameter Network
(incomplete!)
• huge diversity of methods and tools to
determine stellar parameters
• choice of methods and tools depends on:
– precision requirements
– available resources
– observational data available
Stellar distance
©Perryman 2011, fig. 8.1
• Accuracy of distances of planet host stars increased substantially
from ground-based (a) to Hipparcos (b).
• Hipparcos provides proper motions for a large sample of stars and is
a major step in the study of stellar streams.
• … more to come with GAIA!
HR diagram with planet host stars
• HR diagram
of stars
within 25 pc
• based on
Hipparcos
data
•
planet host stars
Hawley & Reid (2003), Perryman (2011; fig. 8.5)
Host stars of planets
•
•
•
•
•
•
•
main-sequence stars
low-mass objects: M dwarfs, brown dwarfs
pulsating stars
giant stars
pulsars
binary and single stars
stellar populations: thick disk, metal-poor
stars, open clusters
Improving precision…
Baines et al. (2008)
• Angular diameters can be measured for nearby stars,
give precise linear radii with Hipparcos distance
… but usually …
• … direct measurement not possible!
• Often, mass and radius are estimated from
evolutionary models.
• Models predict radii and effective
temperature at given mass and age.
• Comparison to observed effective
temperature then gives constraints on mass
and age.
• However, models to be used with care!
Effective temperature
• Luminosity of black body with surface
4πR2 and temperature T:
• Consider star with same luminosity and
same radiating surface area, defines
effective temperature:
radiative flux
Gray (2005), pp. 3, 118ff.
High-resolution spectroscopy
• example: Hα profile of
host stars of planets to
derive effective
temperature precisely
• further parameters can
be derived:
– surface gravity
– chemical abundances
– projected rotational
velocity
– microturbulence
– macroturbulence
Fuhrmann et al. (1998)
How to model a spectrum?
Set of stellar
parameters:
Teff, logg, [M/H],...
Model atmospheres:
e.g. ATLAS9, MARCS, MAFAGS
Temperature profile
Pressure profile
...
Line formation codes:
e.g. SPECTRUM,
MOOG, LINFOR
Synthetic spectrum
Metallicity and chemical abundances
Metallicity and chemical abundances
Astronomer‘s
Metals
More Metals !
Even more Metals !!
The „Bracket“ [Fe/H]
𝑋 𝐻 = log10
𝑁𝑋
𝑁𝐻
𝑁𝑋
𝑁𝐻
⦿
• e.g. [Fe/H] = –1 → 1/10 the iron abundance of the sun
• unit: „dex“ (contraction of decimal exponent, indicates
decimal logarithmic ratio which is in fact unitless)
• [Fe/H] is often used as an overall metallicity indicator,
other elements then are related to Fe, e.g. [Mg/Fe].
III. EFFECTS OF METALLICITY
Metallicity correlation
• host stars of
planets appear on
average more
metal-rich than
comparison stars
without planets
• probability to find
planet is higher
for metal-rich star
Santos et al. (2005)
Is it true?
• selection effects:
– metal-rich stars show deeper absorption lines so that
planets can be detected more easily by RV surveys
– metal-rich stars are intrinsically brighter than metalpoor stars at same spectral type, so that more metalrich stars are selected in magnitude-limited samples
– possibly correlation of orbital radius and metallicity
• Selection effects cannot fully explain the
metallicity correlation!
Metallicity correlation
• So far giant planets around FGK type stars!
• Dependence on nature of star:
– giant stars: no correlation found!
– metal-poor stars: more planets in thick disk (αenhancement!) than in thin disk
• Correlation only for Jovian planets, not for
lower-mass Neptune mass planets!
M dwarfs with planets
Johnson et al. (2010)
• M-type planet host stars tend to be more metal-rich
• but generally less Jovian planets around M dwarfs than
around FGK dwarfs (effect of stellar mass!)
Hypotheses
• primordial origin:
– planet formation preferred in disks of metal-rich
stars
• self-enrichment:
– stellar atmospheres enriched with infalling
planetary material or planetesimals
Primordial origin
• metallicity correlation can be reproduced with
models of core accretion!
• models also predict lower frequency of shortperiod giant planets around M dwarfs and
presence of Neptune-mass ice-giants around
M dwarfs
Self-enrichment
• spectroscopy measures „surface“ abundances
of chemical species!
• enrichment of surface region with:
– engulfment of planet migrating inward
– accretion of disk material resulting from migration
– infall of planetary material or planetesimals
Self-enrichment
Ford et al. (1999)
• accreted material is mixed in outer convection zone
• less atmospheric metal-enhancement when convection
zone is deep
Self-enrichment of giant stars
• planet-bearing stars
with deep convection
zones (sub-giants)
should have less
metallicity
• This is not the case!
Fischer & Valenti (2005)
o sub-giants
Conclusions
• observational evidence for metallicitycorrelation
• probably primordial
• but: accretion of planetary material and
planetesimals is inevitable (Sun!)
Galactic origin
• Galactic radial
metallicity gradient
(0.07-0.1 dex kpc-1)
• radial mixing: most
metal-rich stars
migrated from inner
Galactic disk to solar
Galactocentric
radius
Wielen (1996)
Galactic origin
• planetary systems form in inner metal-rich
disk (independent of metallcity there!)
• also Sun is more metal-rich than local average
and might have formed at inner Galactic radii
• model: metallicity-correlation from radial
mixing of different Galactic components
Model of metallicity-correlation
Haywood (2009)
• frequency of giant planets: metal-poor stars:
0%; local stars: 5%; metal-rich stars: 25%
Conclusions
• metallicity range of planet-bearing stars
corresponds to Galactic ring of molecular
hydrogen linked to star formation
• low-mass planets might simply form in less
dense hydrogen regions
• giant stars with giant planets are relative
young and thus not affected by radial mixing
Condensation
• important for planet formation!
• refractory species:
– high condensation temperature
– planet hosts: Al, Ca, Ti, V
• volatile species:
– low condensation temperature
– planet hosts: C, Cu, N, Na, O, S, Zn
• intermediate: Co, Fe, Mg, Ni
Condensation
• primordial origin of metallicity-correlation:
– expect similar occurence trends for metals other
than iron
• self-enrichment:
– expect overabundance of refractory elements
Refractory Elements
o stars with planets x comparison sample
Bodaghee et al. (2003)
• no abundance trends beyond [Fe/H]
Volatile Elements
Ecuvillon et al. (2006)
• no abundance trend beyond [Fe/H]
III. STELLAR ACTIVITY
Magnetic braking
formation of
convective envelope
Gray (2005), p. 485
Magnetic braking
• convection + rotation are thought to
generate magnetic field via stellar dynamo
(Gray, 2005, pp. 490-492)
Magnetic braking
• stars with convective envelopes form a
magnetic field
• stellar wind is coupled to magnetic field lines
and thus to stellar rotation
• therefore, stellar wind takes away angular
momentum and the stellar rotation is braked
Strassmeier (1997), pp. 68-70
Gray (2005), pp. 492
Magnetic activity
• Active stars show magnetic phenomena
• Stellar dynamos are thought to produce
magnetic fields
• Ionised stellar material couples to magnetic
field lines
• This produces a plethora of phenomena of
magnetic activity: photospheric spots,
chromospheric faculae, coronal holes,
loops, mass ejections, ...
Chromospheric activity
spots
Sun in Ca II K:
chromosphere!
plages
Sun in white light: photosphere
Schrijver & Zwaan (2000), pp. 2,3
chromospheric
network
Measurement principle
famous: Mt. Wilson S-index, here R‘HK index
• measure flux f50 in Willstrop‘s
Active star
band (3925-3975Å; Willstrop,
1964, Mem. RAS 69, 83) for
active star and inactive
standard
• measure angular diameter of
Inactive star
3925Å
standard to get absolute flux
F50 of standard
3975Å
Strassmeier (1997), pp. 249, 250
Measurement principle
•measure flux in emission line
f(H) and f(K)
Active star
Inactive star
•absolute flux in H and K of
active star:
•subtract photospheric
contribution Fphot based on
model atmosphere or inactive
standard:
Strassmeier (1997), pp. 249, 250
Measurement principle
Active star
• absolute flux in H and K of
active star:
Inactive star
Strassmeier (1997), pp. 249, 250
Activity-related RV
Saar et al. (1998, ApJ 498L, 153)
Activity-related RV
largest RV scatter:
•active F stars
•dMe stars
•high Ca II H&K emission
Saar et al. (1998, ApJ 498L, 153)
Activity-related RV
• RV scatter scales with rotational speed!
• Young G-type (0.3 Gyr) star with vsini=8-10km/s: 20-45 m/s
Saar et al. (1998, ApJ 498L, 153)
• Good agreement with expectations from convective
motions and spots!
Solar and stellar flares
• flares may have dramatic consequences for
life on planets
• solar outer atmospheric layers: 10-4 of
photospheric radiation on average
• local, short-lived explosive events may
exceed average by factor 103-104, i.e. larger
than photosperic flux!
Stix (1991, pp. 351-357)
Solar and stellar
flares
pre-flare
post-flare
disk view:
disappearing
filament
limb view: eruptive promincence
maximum
expanding
ribbons
Stix (1991, pp. 352, 355)
Solar and
stellar flares
Stix (1991, p. 353)
thermal
thermal
non-thermal
Solar and
stellar flares
• thermal regime: up to approx. 107K
• non-thermal regime:
bremsstrahlung and synchrotron
radiation of electrons with 10-100
keV
Stix (1991, p. 353)
Solar and stellar flares
• solar flares: up to 1031erg, brightening not
visible in disk-integrated light
• stellar case: only disk-integrated light
measurable!
• stellar flares: large flares which are also
visible in white light
• super flares: 1033-1038erg
• Sun: no super flares during last 2000 years
Stix (1991, p. 354)
Schaefer et al. (2000, ApJ 529, 1026)
Stellar super flares
Schaefer et al. (2000, ApJ 529, 1026)
Stellar super flares
•
•
close to main-sequence
spectral types F8-G8
• isolated
• slow or moderate rotators
Schaefer et al. (2000, ApJ 529, 1026)
Stellar super flares and effects on
planet
• energy deposited on a planet at 1 AU by a
super flare with 1035erg: 3.5x107 erg cm-2
• no melting of rock but of ice
• planet with atmosphere:
– possibly temporary heating, ozone depletion
– formation of organic molecules (Miller-Urey
experiment)
Schaefer et al. (2000, ApJ 529, 1026)