What can transit observations tell
us about (exo)-planetary science?
Part II – “Spectroscopy” & Atmospheric
Composition/Dynamics
Kudos to Heather Knutson,
now at Caltech!
Ge/Ay133
Rapid Progress: Transiting Planets, 1 May 2007
One year later (2008): 43 Systems And Counting
Ice/Rock Planets
Updates from exoplanets.org :
Hot Jupiter/Neptune
atmospheres?
M
L
T
In the optical/near-IR,
the spectra of M → T
dwarfs (similar temp.
as the hot Jupiters)
show strong alkali
metal lines:
Transiting Planets as a Tool for Studying
Exoplanet Atmospheres
Secondary Eclipse
See thermal radiation and
reflected light from planet
disappear and
reappear
Transit
See radiation from star
transmitted through the
planet’s atmosphere
Orbital Phase Variations
See cyclical variations in
brightness of planet
Characterizing Atmospheres With
Transmission Spectroscopy
Star
Planet
Atmosphere
• Probes composition of atmosphere
at day-night terminator
• Can search for clouds, hazes,
condensates
HST STIS transits of HD 209458b from
290-1030 nm (Knutson et al. 2007a)
First detection of
an extrasolar planet
atmosphere:
Look for the
transit depth in
filters on and
off the Na Dline with HST.
Charbonneau, D. et al. 2001, ApJ, 568, 377
Vidal-Madjar, A. et al. 2004, ApJ, 604, L69
Atmospheres Part II:
Most atoms have their
so called resonance
lines in the UV. The H I
depth is VERY large.
EXOSPHERE?
Water and Haze on HD 189733b
Featureless visible light
spectrum indicates hazes…
Figure from Pont, Knutson et al. (2007) showing
atmospheric transmission function derived from HST
ACS measurements between 600-1000 nm
… which disappear in infrared,
revealing water absorption features.
Figure from Swain et al (2008) showing infrared atmospheric
transmission function derived from HST NICMOS spectra compared
to models for the planet’s transmission spectrum with (orange) and
without (blue) additional methane absorption (Tinetti et al. 2008).
What about day/night chemistry? Need IR observations:
GL 229B (BD)
Oppenheimer, B. et al. 1998, ApJ, 502, 932
T dwarf IR opacities dominated by CH4, H2O.
Use secondary eclipses to acquire dayside fluxes:
A Broadband Emission Spectrum For
HD 189733b
Charbonneau, Knutson et al. (2008), Barman (2008)
Can even collect R~50-100 spectra: IRS Data for HD 189733b
Grillmair et al., Nature 456, 767
(Dec. 11 2008)
Gillett, Low, & Stein (1969), “The 2.8-14
Micron Spectrum of Jupiter”
“Most of the features of the 2.8-14 μm
spectrum of Jupiter can be accounted for
on the basis of absorption by NH3, CH4,
and H2.”
A Near-IR Emission Spectrum for HD 189733b
HST NICMOS observations
of a secondary eclipse of HD
189733b
“Most of the features of the 2.8-14 μm
spectrum of Jupiter can be accounted for
on the basis of absorption by NH3, CH4,
and H2.”
Swain et al. (2009)
Even in space, these measurements are at the limits of current detectors:
HST
NICMOS
Spitzer
A Surprise: The Emission Spectrum of HD 209458b
Requires a model with a
temperature inversion and
water features in emission
instead of absorption.
Knutson et al. (2008c),
Burrows et al. (2007)
Why would two hot Jupiters with similar
masses, radii, compositions, and
temperatures have such different pressuretemperature profiles?
Gas Phase TiO/VO
Temperature Inversion?
Problem: Cold Trap
TrES-4 is a great
test case!
Teq = 1760 K
One alternative:
photochemistry
(tholins, polyacetylenes?)
Inverted
Non-Inverted
As described in Hubeny et al.
(2003), Burrows et al. (2007,
2008), and Fortney et al. (2008)
Figure from Fortney et al. (2008)
Increasing UV
Possible Explanation: UV Chromospheric Stellar Activity?
Figure from Knutson et al. 2010, ApJ, 720, 1569
Ultimately want many objects/wavelengths;
Problem: Switch from inverted to non-inverted
states can artificially increase day-night contrast
Model for HD 209458b from
Showman et al. (2008)
Solution: Use 3.6 and 4.5 μm phase curves
to map extent of inversion
Observations of HD 209458b
from Knutson et al. (2008a)
Ultimately want many objects/wavelengths;
Solution: Use 3.6 and 4.5 μm phase curves
to map extent of inversion
Test of Spitzer color index with stellar
UV activity Knutson et al. 2010, ApJ,
720, 1569
Observations of HD 209458b
from Knutson et al. (2008a)
The warm Spitzer mission has done another
18 planets at 3.6/4.5 mm (H. Knutson, P.I.).
Ground? Line shape would give pressure at the photosphere,
center/shift the wind profiles. Challenge is the Earth’s atmosphere!
Terrestrial CH4
CO Search
Limits only just beginning to
reach sufficient sensitivity…
Deming, D. et al. 2005, ApJ, 622, 1149
First possible ground based high spectral resolution detection:
Gives orbital
velocity and
thus absolute
mass of the
planet & star
(w/RV), is the
blueshift due
to strong
winds across
the terminator?
Snellen, I. et al. 2010, Nature, 465, 1049
A Diversity
of Worlds
Super-Earths & MiniNeptunes
Mass range:
1-10 Earth masses
Prospects for Studies of Terrestrial Planets With the James
Webb Space Telescope (launches 2018?)
Predicted transmission spectrum for a 0.5 Mearth,
1 Rearth, 300 K planet orbiting a M3V, J=8 star
Neptune-mass planets are
observable with Spitzer and
HST….
… but observations of earthlike planets orbiting M
dwarfs will require JWSTlevel precision
Seager, Deming, & Valenti (2008)
Imaging extrasolar planetary systems?
Marois et al. (2008),
Science
Jovian-mass planets cool
slowly, so few-few 10s of MYr
old objects are fairly bright…
And have emission peaks in the
near-IR atmospheric windows where
AO systems perform well.
Signatures of “young” planetary systems?
HR 8799
Marois et al. (2008), Science
One group of systems to try are the
so-called ‘debris disks’ that we’ll
learn about later. These are young
stars with “2nd generation dust”
caused by planetesimal collisions.
 Pictoris (VLT,
proper motion
now confirmed)
Can also use coronography/PSF subtraction in space:
Too bright @
600 nm?
Circumplanetary disk?
If so, Mp?
Use dynamics!
Fomalhaut dynamical analysis of
companion mass:
Kalas, P. et al. (2008), Science
Neptune-mass planets are
observable with Spitzer and
HST….
Modeling of the dust ring
suggests an upper limit to the
companion of ~3 MJ.
Photometry-based mass
estimate uncertainties are
dominated by possible age(s).
Formation? (In situ/scattering?)