Substellar Atmospheres IV.
Weather at the L/T Transition
PHY 688, Lecture 21
Mar 16, 2009
Outline
•
Course administration
– oral presentation sign-up: April 27– May 8
– problem set 3 now on-line; due Monday, Mar 23
– seminar by Adam Burrows: Wed 1pm in ESS 450
•
Discussion of problem set 2
– problem 1: curve of growth
– problem 3: extrasolar planet
– problem 4: open and globular cluster metallicities
•
Review of previous lecture
– clouds levels and emergent flux from L and T atmospheres
– current astrophysics problem: rapidity of cloud disappearance at the L/T transition
•
Modeling the L/T transition: weather
– possible scenarios
– observational constraints
– relevance to extrasolar planets
Mar 16, 2009
PHY 688, Lecture 21
2
Problem 1: Curve of Growth
re
a
squ t
roo
1
"W #
log $
%
&
!
'
(
flat
0
1
r
a
e
lin
!
!
1
W" N
W " ln N
W "N
0
1
!
2
3
4
log (Nf )
• in our case: a = γ / (2 ∆λ)
– i.e. Lorentzian / Doppler width
• flat part of c.o.g. seems missing for large a
Mar 16, 2009
PHY 688, Lecture 21
3
From Lecture 12: Lorentzian vs.
Gaussian Line Profiles: Large τ
simulation
for the Hα
line profile
• core more
sensitive to
Gaussian
parts
• wings more
influenced
by Lorentzian parts
Mar 16, 2009
PHY 688, Lecture 21
4
Outline
•
Course administration
– oral presentation sign-up: April 27– May 8
– problem set 3 now on-line; due Monday, Mar 23
– seminar by Adam Burrows: Wed 1pm in ESS 450
•
Discussion of problem set 2
– problem 1: curve of growth
– problem 3: extrasolar planet
– problem 4: open and globular cluster metallicities
•
Review of previous lecture
– clouds levels and emergent flux from L and T atmospheres
– current astrophysics problem: rapidity of cloud disappearance at the L/T transition
•
Modeling the L/T transition: weather
– possible scenarios
– observational constraints
– relevance to extrasolar planets
Mar 16, 2009
PHY 688, Lecture 21
5
Previously in PHY 688…
Mar 16, 2009
PHY 688, Lecture 21
6
Direct Evidence for
Dust in L Dwarfs
•
•
Model photospheres are
excellent at reproducing the
spectra of L and T dwarfs in the
mid-IR L (unlike in the optical
and in the mid-IR)
However, 9–11µm “plateau” in
mid-L dwarfs was not
anticipated a priori
– flux deficiency
•
Likely the result of the direct
detection of electronic
transitions in silicates
– Si-O stretching mode
Mar 16, 2009
PHY 688, Lecture 21
(Cushing et al. 2006)7
Emergent Flux Depends on Wavelength
and Cloud Level
τcloud < 0.5; hcloud > hphotosphere
τcloud > 1; hcloud ~ hphotosphere
silicate cloud
(frain = 3)
τcloud > 1; hcloud < hphotosphere
Mar 16, 2009
PHY 688, Lecture 21
(Ackerman & Marley 2001) 8
From Lecture 8:
Near-IR CMD of Stars and Brown Dwarfs
F–K
M
L
•
dusty clouds are the reason that
L to early-T dwarfs are
unusually red
•
sedimentation (rain-out) of
clouds in mid-T dwarfs
removes clouds as a source of
opacity and reddening
•
… but why should T dwarfs be
bluer than M dwarfs in the
near-IR?
T
Mar 16, 2009
PHY 688, Lecture 21
9
Cloud Sedimentation Is Coincident
with CO to CH4 Transition
Mar 16, 2009
PHY 688, Lecture 21
(Burrows et al. 2001)
10
Modeling
L and T Dwarfs
•
•
•
Models that incorporate
suspended dust (DUSTY)
successfully reproduce L
dwarf colors
M
L
T
Late T dwarfs well fit by
dust-free photospheres (e.g.,
COND models: dust removed
upon formation)
DUSTY models
(dust remains
suspended)
COND models
(dust is removed)
Transition can be explained
by sedimentation of silicate
clouds below visible
photosphere
Mar 16, 2009
PHY 688, Lecture 21
11
(Baraffe et al. 2003)
The L/T Transition Problem
•
•
photospheres turn blue in the
near-IR unusually quickly
clouds sink comparatively
slowly
fr
ai
n
→
∞
– need to be “rained out”
(sedimented) faster
3
=
n
f rai
Mar 16, 2009
PHY 688, Lecture 21
(Burgasser et al. 2002)
12
From Lecture 8:
Effective Temperatures of Brown Dwarfs
•
rather than cooling between
late-L and mid-T spectral
types, brown dwarfs merely
change appearance
– change in temperature is
∆Teff < 150 K
•
models of constant frain
require ∆Teff ~ 600 K of
cooling to reproduce the
same effect on the emergent
SED
Mar 16, 2009
PHY 688, Lecture 21
13
(Kirkpatrick 2005)
The L/T Transition Problem
•
•
photospheres turn blue in the
near-IR unusually quickly
clouds sink comparatively
slowly
– need to be “rained out”
(sedimented) faster
Mar 16, 2009
→
ai
n
fr
reddest L dwarfs require
inefficient sedimentation
(frain < 3)
• early T dwarfs require frain > 3
• late T’s require no visible
clouds (frain → ∞)
∞
•
3
=
n
f rai
PHY 688, Lecture 21
(Burgasser et al. 2002)
14
Another L/T Problem: J-band Flux
Reversal in Early- to Mid-T Dwarfs
10 m Keck telescope laser-guide star adaptive optics images
T1
0.1336"
T5
• 2MASS 1404 AB: T1 + T5 dwarf binary
• the earlier-type T dwarf, which is bolometrically
brighter, is fainter at J band (only)
Mar 16, 2009
PHY 688, Lecture 21
(Looper et al. 2008)
15
The J-band Flux
Reversal in Earlyto Mid-T Dwarfs
Mar 16, 2009
PHY 688, Lecture 21
(Looper et al. 2008)
16
What Is the Weather on
an Early T Dwarf?
• partly cloudy?
• uniformly hazy?
• raining “cats and dogs”?
– i.e., silicates and iron
Mar 16, 2009
PHY 688, Lecture 21
17
Outline
•
Course administration
– oral presentation sign-up: April 27– May 8
– problem set 3 now on-line; due Monday, Mar 23
– seminar by Adam Burrows: Wed 1pm in ESS 450
•
Discussion of problem set 2
– problem 1: curve of growth
– problem 3: extrasolar planet
– problem 4: open and globular cluster metallicities
•
Review of previous lecture
– clouds levels and emergent flux from L and T atmospheres
– current astrophysics problem: rapidity of cloud disappearance at the L/T transition
•
Modeling the L/T transition: weather
– possible scenarios
– observational constraints
– relevance to extrasolar planets
Mar 16, 2009
PHY 688, Lecture 21
18
Scenario 1: Disruption of Clouds
•
•
•
i.e., “partly cloudy” model
all L dwarfs cool through the
end of the L spectral type
sequence
at ~1300 K something starts
rapidly disrupting the clouds
– possibly because of
sufficiently deep settling into
convection zone, so that they
become subject to global
circulation pattern
•
J-band brightening: due to
appearance of hotter layers at
1–1.2 microns
Mar 16, 2009
PHY 688, Lecture 21
(Burgasser et al. 2002)
19
From Lecture 20: Emergent Flux
Depends on Wavelength and Cloud Level
τcloud < 0.5; hcloud > hphotosphere
τcloud > 1; hcloud ~ hphotosphere
silicate cloud
(frain = 3)
τcloud > 1; hcloud < hphotosphere
Mar 16, 2009
PHY 688, Lecture 21
(Ackerman & Marley 2001) 20
Scenario 1: Disruption of Clouds
•
•
•
i.e., “partly cloudy” model
all L dwarfs cool through the end
of the L spectral type sequence
at ~1300 K clouds start rapidly
disrupting
– e.g., because of settling
sufficiently deep into the
convection zone, so that they
become subject to global
circulation pattern
•
•
•
J-band brightening: due to
appearance of hotter layers at
1–1.2 microns
need for new parameter:
fraction of cloud cover
based on idea by Marley et al.
(2002)
Mar 16, 2009
PHY 688, Lecture 21
(Burgasser et al. 2002)
21
1. Disruption of Clouds: Predictions
•
•
would not expect many
early T’s
possible photometric
variability due to patchy
cloud cover
– hotter regions visible
through holes in clouds
•
possible re-appearance in
early T’s of 1–1.2 micron
spectroscopic signatures
characteristic of hotter midL dwarfs (e.g., FeH, CrH)
Mar 16, 2009
PHY 688, Lecture 21
(Burgasser et al. 2002)
22
Scenario 2: Unified Cloudy Models
•
•
•
no disruption of clouds, just regular settling
lower-mass (-gravity) brown dwarfs start turning to the blue at brighter
magnitudes
attractive because does not increase model complexity
Mar
16, 2009
(Tsuji
& Nakajima 2003)
PHY 688, Lecture 21
23
2. Unified Cloudy Models: Predictions
•
•
•
•
spread in MJ vs. J – K, dependent on gravity
lower-mass transition dwarfs should be brighter
J-band brightening is an artifact of the spread in gravities/ages
~ uniform population of MJ vs. J – K phase space around the L/T transition
Mar
16, 2009
(Tsuji
& Nakajima 2003)
PHY 688, Lecture 21
24
Scenario 3: Sudden Downpour
•
•
•
rain causes rapid
cloud thinning at
~1400 K
also requires extra
parameterization:
rate of change of
fsed (≡ frain)
favored by Saumon
& Marley (2008)
Mar 16, 2009
(Knapp et al. 2004)
PHY 688, Lecture 21
25
3. Sudden Downpour: Predictions
• J-band
brightening
• likely no
variability
– uniform thinning
of clouds
• few early T’s
Mar 16, 2009
(Knapp et al. 2004)
PHY 688, Lecture 21
26
The Observational Evidence:
I. J-band Flux Reversal
10 m Keck telescope laser-guide star adaptive optics images
T1
0.1336"
T5
•
•
•
•
2MASS 1404 AB: T1 + T5 dwarf binary
the earlier-type T dwarf, which is bolometrically brighter, is fainter at J band
(only)
this is recent evidence that post-dates the uniform cloudy model (Tsuji &
Nakajima 2003, scenario 2)
very difficult to explain with UCM model
Mar 16, 2009
PHY 688, Lecture 21
(Looper et al. 2008)
27
The Observational Evidence:
II. Spectral Type Distribution of T Dwarfs
•
•
•
~ 1/3 of T dwarfs are early T’s
but because of being more luminous, early T dwarfs are much rarer in actual fact
relative dearth of early T’s agrees with patchy cloud and sudden downpour
scenarios (1. and 3.)
Mar 16, 2009
PHY 688, Lecture 21
(Metchev et al. 2008)
28
The Observational Evidence:
III. Photometric Variability in Jupiter
bright 5-micron regions
reveal deeper, hotter layers
false-color near-infrared
Mar 16, 2009
PHY 688, Lecture 21
5 micron
29
The Observational Evidence:
III. Photometric Variability in Jupiter
5 micron
•
Jupiter’s 5-micron “hot spots”
– visible in cloud holes, cover small fraction of surface, incur measurable variability
•
similar ~1-micron hot spots in brown dwarfs? Evidence is inconclusive
Mar 16, 2009
PHY 688, Lecture 21
(Gelino & Marley 2000)
30
The Observational Evidence:
IV. Re-appearance of FeH in
Early- to Mid-T’s
• FeH characteristic of hotter
temperatures (>1400 K) of L
dwarfs
• re-appearance in T’s best
explained by patchy cloud
cover
– FeH visible through holes in
clouds
Mar 16, 2009
PHY 688, Lecture 21
(Burgasser et al. 2002)
31
Relevance to Planets
• The only extrasolar planets currently imaged appear to be
near the L/T transition
2MASS 1207 A/B
VLT AO image
B
5–8 MJup
0.78” (55 AU)
(Chauvin et al. 2004)
Keck adaptive optics image of the
HR 8799 bcd planetary system;
near-IR color composite
(Marois et al. 2008)
Mar 16, 2009
PHY 688, Lecture 21
32
Relevance to
Planets
• The only
extrasolar planets
currently imaged
appear to be near
the L/T transition
– HD 8799 bcd
– 2MASS 1207 B
Mar 16, 2009
PHY 688, Lecture 21
(Marois et al. 2008)33
Relevance to Planets
• The only extrasolar planets currently imaged appear to be
near the L/T transition
• Inferred temperatures of transiting hot Jupiters are also
near 1000–1500 K
– nominally close to the ~1400 K L/T transition temperature
Mar 16, 2009
PHY 688, Lecture 21
34
Young and Cool Brown Dwarfs:
the Domain of Hot Jupiters
Mean field age: ~3–5 Gyr
log g ~ 5.5
(~300 Myr)
log g ~ 3.0–5.0
(~300 Myr)
(~10 Myr)
Metchev & Hillenbrand (2006); Luhman et al. (2007); Mohanty et al. (2007)
Mar 16, 2009
PHY 688, Lecture 21
35
Relevance to Planets
• The only extrasolar planets currently imaged appear to be near the
L/T transition
• Inferred temperatures of transiting hot Jupiters are also near
1000–1500 K
– nominally close to the ~1400K L/T transition temperature
• Presence of clouds governs atmospheric albedo
– e.g., Jupiter, Venus
– higher albedos promise easier
detection of extrasolar planets
in reflected light
Mar 16, 2009
PHY 688, Lecture 21
36
Relevance to Planets
• The only extrasolar planets currently imaged appear to be near the
L/T transition
• Inferred temperatures of transiting hot Jupiters are also near
1000–1500 K
– nominally close to the ~1400K L/T transition temperature
• Presence of clouds governs atmospheric albedo
– e.g., Jupiter, Venus
– higher albedos promise easier
detection of extrasolar planets
in reflected light
Mar 16, 2009
PHY 688, Lecture 21
37