Transcript So, what`s the problem for high

We’ve got the smoke, but where’s the gun?
IR astronomy, paradigms of star formation,
and
A search for high-mass protostars.
Murray Campbell, 25 years of Colby students, and colleagues at UofA,
UT Austin, BU, CfA, Keck, UKIRT, Cornell, Gemini, and MPIfA
Current students at Colby:
Frank Fung, Tomas Vorobjov, Cliff Johnson, and Ry Brooks
Special thanks to Mike Ramstrom, and John Kuehne,
?
Only on the
largest
cosmological
scales is
astronomy
simple!
The
complexity
you see here is
only the
beginning…
A new composite
view of the
Orion nebula.
More beauty.
More
complexity.
Hubble:Advanced Camera for Surveys Ground ESO-MPI 2.2m La
Silla Filters:F435W (B) F555W (V) F658N (Ha) F775W (i)
F850LP(z)ESO842 (B) ESO856 (Ha) ESO857 ([S II]) ESO859 ([O
III])
The Simple Analysis Party is Busted!



Astronomy ought to be
a study of atoms,
gravity, nuclear physics,
and newsworthy exotic
forces.
Dust, clumpiness, and
turbulence spoil the
party.
The universe is 90% H,
and clouds are mostly
in H or H2, but you can’t
see them in visible light.
Interstellar H or H2 cloud
in visible light:
Shit Happens.
Orion is a rare H+ = HII cloud
(HII region)
Redeeming graces of Dust

Wavelength shifter
and integrator:
Absorbed light is reemitted in IR.
IR is often THE best
estimator of
luminosity!
Great Observatories Origins Deep Survey
Figure 24-36
Dusty Donut
Chaisson and McMillan,
Astronomy Today, 5th ed.
(2005)
The Virial Theorem Rules Many
Processes, including Star
Formation
1 2
1
 2mv  2 Gravitational Potential Energy
Gas,etc




As an object collapses
to a smaller size, the
magnitude of its Gravitational Potential Energy gets
bigger.
But only half of the kinetic energy released can go
into heating the gas, or it violates the theorem.
If it can’t radiate, it can’t collapse!
Dust is a key player: I
and H2 don’t radiate in most cloud
conditions.
 Dust can absorb energy from gas through
collisions and radiate it in the IR.
 Dust’s IR efficiency can keep the cloud at
its ORIGINAL TEMPERATURE.
 So the collapse is initially FREE FALL!
 Density variations cause fragmentation
into many cores.
H
Dust is a key player: II

Each collapse is inside out: a collapsing core
becomes a protostar inside an envelope.
 The outside envelope’s dust absorbs the
protostar’s emission and reradiates it in the
IR.
 So we can find where the process is going
on.
 But the protostar itself may be obscured.
Shit Happens.
Dust is a key player: III
 In
far-IR (50-200m), dust emission gives
virtually total luminosity (power in watts).
 In millimeter and submillimeter
wavelengths, dust emission can be used
to measure the mass of a molecular cloud,
and its star forming cores.
Oops.
Galaxies, Dust, and
Starbursts
I forgot to say that “naturally
occurring” interstellar clouds are
stable against collapse.
The density is too low to have
enough gravitational potential
energy and the collisional
coupling between gas and dust is
too weak for dust cooling.
They only become unstable with
help from a compression, as
happens in galactic collisions.
Hubble close-up of a galactic
collision that triggered star
formation in NGC1275. The star
formation is in the optically dark
clouds.
A Brief History




The first observation of dust in
emission was made by
physicists who built IR
detectors and put them on
telescopes.
The first important far-ir source
was the Galactic Center,
discovered by a one-inch
telescope on a high altitude
balloon.
It’s luminosity comes largely
from formation of high-mass
stars.
The Infrared Astronomy
Satellite (IRAS) discovered
that many galaxies are strong
IR emitters, especially colliding
galaxies!
The Galactic Center and plane
of the the Milky Way at 100
m are shown in the
background, but Spitzer is
best for galaxies and lowmass stars.
More History…

The bright far-IR
sources were all dust
clouds around highmass stars.
 Radio astronomers
mapping molecules’
rotation lines
discovered incredible
outflows!
 The outflows were
associated with lowmass protostars.
Most astronomers focused on
low-mass star formation.
Low-Mass Star Formation: Paradigms for the
cloud core’s evolution.

The core is the most
easily observed
component
 Before collapse, the
core has uniform
temperature, and
density ~r-2, or
something else.
 A uniform core would
experience a
homologous collapse.
Dense
parts collapse
fastest.
A cloud forms many
cores.
In each core, the
protostar pulls away
from the outer core:
inside out-collapse.
Rotation is
important…
A (Simplified) Realistic Model

An outer envelope,
with density ~r-2.
 A rotationally flattened
infalling envelope with
density determined by
free fall, ~r-3/2.
 An inner, flared
accretion disk.
 An outflow cavity.
Barbara Whitney et al. 2003 ApJ
591, 1049
WHY SHOULD WE BELIEVE THE PARADIGM?




Hydrodynamic models of
collapsing clouds are roughly
fit by observations.
Hydrodynamic models of
protostars give reasonable
luminosities due to
gravitational contraction,
accretion, and the onset of
nuclear fusion.
Hydrodynamic models of
accreting protostars are
consistent with reasonable
accretion rates.
Observed Spectral Energy
Distributions (SEDs) can be
explained by evolving disks
and envelopes.
Jets are a natural consequence of
rotation, even if hard to understand
in detail.
 Disks are observed.

Interstellar Cloud Evolution Shown Explains a “Sequence” of SEDs
Protostar Class 0/I
Class II/III
Star
Chaisson and McMillan (2005)
If a protostar’s a gun, what forms does
the smoke take?

Dust





cm continuum
mid-ir ionic lines [NeII]






Near-ir lines from shock
fronts

mm emission lines
Velocity maps of outflows
Velocity maps of disks
Molecules--non-thermal
(Masers)

Molecular hydrogen, H2

Molecules--thermal

Ionized gas (HII regions)
around young high-mass
stars


Far-ir continuum emission
Mid-ir cont. emission and
solid state bands
mm and sub-mm cont.
emission

H 2O
CH3OH (methanol)
SiO, OH, etc.
Visible images of disks
and jets around low-mass
protostars
Embedded low-mass
young stars (YSOs)
visible in near-ir
So, what’s the problem for
high-mass star formation?








Very high accretion rates are required.
Nuclear fusion should begin in the core before the outer layer is
finished accretion from its parent cloud core.
Radiation pressure should stop accretion before a star can reach its
final mass.
High-mass stars only form in clusters, so isolating individuals is
difficult: Almost no HMPOs have been unambiguously identified at
specific star-like points on the sky that can be easily observed in
isolation from nearby already formed high-mass stars.
High-mass stars are rare, so high-mass protostars (HMPOs) are
even less common, especially nearby.
HMPOs are more deeply embedded than low-mass counterparts, so
there are no certain visible light images of disks and jets.
Obscuration by clumps of dense gas and dust make identification of
individual protostars difficult.
No details are really pinned down in the process.
Why are HMPOs important?
 High-mass
stars dominate the luminosity
of galaxies they are in.
 High-mass stars’ nucleosynsthesis
dominates the chemical evolution of
galaxies, and the universe.
 High-mass stars luminosities, outflows,
and winds dominate the evolution of the
interstellar medium, and hence on-going
star formation.
What should we look for?
 Sources
of mid- and far-ir with a variety of
sensible spectral energy distributions
(SEDs).
 Hot, high density molecular cores that are
internally heated.
Evolution of SEDs
Very approximate calculations by Cliff Johnson and me for high-mass protostars.
Star, "Disk", and Envelope: Av Disk = 0.05 Av Envelope = 10
1.00E-07
1.00E-07
Lambda F(lambda) in erg sec-1 cm-2
1.00E-06
1.00E-08
1.00E-09
1.00E-10
1.00E-08
1.00E-09
1.00E-10
1.00E-11
1.00E-11
0.1
1
10
100
1000
0.1
1
10
Wavelength in Microns
100
1000
Wavelength in Microns
Star, "Disk", and Envelope: Av Disk = 0.0 Av Envelope = 0
Star, "Disk", and Envelope: Av Disk = 0.01 Av Envelope = 0
1.00E-06
1.00E-06
1.00E-07
Lambda F(lambda) in erg sec-1 cm-2
1.00E-07
Lambda F(lambda) in erg sec-1 cm-2
Lambda F(lambda) in erg sec-1 cm-2
Star, "Disk", and Envelope: Av Disk = 0.5 Av Envelope = 200
1.00E-06
1.00E-08
1.00E-09
1.00E-08
1.00E-09
1.00E-10
1.00E-10
1.00E-11
0.1
1.00E-11
0.1
1
10
Wavelength in Microns
100
1000
1
10
Wavelength in Microns
100
1000
Color-Color Plots

The shapes of SEDs can
be analysed by ratios of
logs of photometric
points.
 Ratios of logs are
proportional to
differences in
astronomical magnitudes,
that are called “Color
Indices”
 “Color-Color” plots can be
used to select classes of
objects from a
photometric database.
Wood and Churchwell derived
color-color criteria for deeply
embedded high-mass stars with
ultracompact HII regions (UC
HIIs).[1989 ApJ, 340, 265)]
Sridharan, Beuther, Schilke and
Menten Survey of HMPOs I
Criteria for sources:
 Detected in previous CS
survey.
 IRAS IR colors meeting Wood
and Churchwell UCHII region
criteria.
 Bright in far-ir in IRAS survey.
(IRAS F(60 m) > 90 Jy and F(100 m) >
500 Jy)

Not detected in cm surveys for
HII regions. (F(cm) < 25 mJy)
Number satisfying criteria
 69
Observations:
 Search for mid-ir sources in
Midcourse Space Expreiment
(MSX) archive.
 Search for cm emission at Very
Large Array.
 Survey thermal molecular lines at
Institut de Radioastronomie
Millimetrique (IRAM) 30 m
telescope and MPIfR Effelsberg
100m telescope .
 Search for masers at MPIfR
Effelsberg 100m telescope.
 Map dust continuum at IRAM.
Sridharan, Beuther, Schilke and
Menten Survey of HMPOs II
Some Results:
 Derived Luminosities from IRAS.
 Found most sources in 18 MSX
survey.
 Found some to be weak cm sources
indicative of young UCHII regions.
 Found many with wide CO lines
indicating outflows.
 Found many H20 and CH30H masers.
 Mapped all in 1.2 mm dust continuum.
Why Follow-up Sridharan
Survey on IRTF?

IRAS and MSX surveys were low resolution, with
beams (1x5 and 18) covering whole protoclusters, so hope to:
 Identify individual HMPOs with beams of 1-2.
(IRTF would primarily detect compact objects)
 Determine statistics of HMPOs in each protocluster, the initial mass function (IMF).
 Make approximate models of dust around each
HMPO.
Observations at IRTF on
Mauna Kea, Hawaii
In collaboration with Shridharan, Beuther, and MIRSI
Team.
• High sensitivity images at 10.4m covering about a
square arc minute.
• Images at 24.8m.
• Low resolution “grism” spectra from 8-13m.
• Telescope was controlled from my office 9/13-15/2003.
• Frank Fung worked on planning and initial data
processing.
• Tomas Vorobjov helped with planning.
• Frank, Glen Munkhold, and my wife kept logs during
parts of the observations.
•
Criteria and Initial Results
 Chose
1/3 of survey members with
brightest, most compact MSX 12m
emission.
 Found compact sources in 18 of 23 fields.
 Found multiple sources in 7 fields.
 On average, brightest source in each field
accounts for 40% of large beam flux.
 Typical spectra have moderately deep
silicate absorption.
Analysis in Progress

Studying two fields that have been mapped on
the Plateau de Bure Interferometer (PdBI) or the
Submillimeter Array (SMA) .
 Reducing and analyzing spectra--begun by Ry
Brooks.
 Beginning to make approximate models of SMA
sources--begun by Cliff Johnson.
IRAS 19410+2336



Galactic Legacy Infrared
Midplane Survey Extroidinaire
PI: Ed Churchwell, Wisconsin
Spitzer Space Telescope
GLIMPSE IRAC Image
4ox4o
Blue=3.6m
Green=5.8m (H2)
Red=8.0 m (PAH)
Collaborator Joe Hora (CfA) has
accessed the Spitzer database, and
processed the IRAC images.
Shocked H2 and CO Outflows
Our region of
interest
Beuther, Schilke&
Stanke
2003A&A 408,601
Our region of
interest
Dust Continuum (1.2mm, 3 mm, 1.3mm) Beuther & Schilke 2004
Science 303, 1167
MM and Mid-IR Compared: Scales alligned
Right: 1.3mm
Plateau de Bure
Interferometer
Blue: IRAC 8m; Green IRTF 10.4m; Red
IRTF 24.8m
IRAC Positions
What is the optical depth of the dust in the mid-ir?






Li & Draine 2001 ApJ 554, 778
Absorption/emissivity properties of
diffuse interstellar dust are
dramatic function of wavelength.
Based on 1.2 mm dust emission,
expect Av = 1000 in core.
Don’t expect to see any mid-ir at
all!
Source might be on near edge of
cloud core--would violate
paradigm for high-mass star
formation in center of core.
Source might be viewed through
an outflow cavity aimed right at us.
Should try imaging disk at shorter
ir (5m)!
Why don’t we see a cluster of sources?
Clump mass function looks like the Initial Mass
Function--Maybe we see only the single brightest star.
Fig. 2. The mass spectrum of
IRAS 19410+2336. The clumpmass bins are [1.7(3),4], [4,6],
[6,8], [8,10], and [10,25] M,
and the axes are in logarithmic
units. The error bars represent
the standard deviation of a
Poisson distribution . The solid
line shows the best fit to the
data N/M M–a, with a = 2.5.
The dashed and dotted lines
present the IMFs derived from
Salpeter with a = 2.35 (15) and
Scalo with a = 2.7 (17),
respectively.
What about all the Shocked H2 and CO Outflows?
Maybe they all go with the lower mass stars.
Maybe the high-mass star has passed the outflow stage:
There is a weak cm (UC HII region) near the brightest mm peak.
But there are also an H2O maser and a CH3OH maser near the brightest mm peak.
Originally observed in partly cloudy skies--to be observed on Gemini.
IRAS 18089-1732

Spitzer Space
Telescope GLIMPSE
IRAC Image
 4ox4o
 Blue=3.6m
Green=5.8m (H2)
Red=8.0 m (PAH)
Galactic Legacy Infrared Midplane
Survey Extroidinaire
PI: Ed Churchwell, Wisconsin
Collaborator Joe Hora (CfA) has
accessed the Spitzer database, and
processed the IRAC images.
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
IRAS 18089-1732 in Sub-mm


Beuther et al.2004 ApJL 616, L23



In mm and sub-mm
dust continuum it’s
only one source!
UC HII and H2O
maser are at center.
CH3OH maser is
offset from center.
There is a jet in SiO.
There is a disk in
HCOOOCH3.
Beuther et al. 2005 ApJ 628, 800. * is dust peak position.
Four Mid-IR Views of IRAS
18089-1732
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
The action’s here!
IRAS 18089-1732 in Sub-mm
A
B
Beuther et al.2004 ApJL 616, L23
Beuther et al. 2005 ApJ 628, 800.
* is dust peak position.
Cool Image!
The image at source B clearly
indicates a temperature gradient!
But it does not fit well into any
simple symmetry.
And it does not indicate where
the protostar is.
Shit Happens!
A
B
Interpretation:
 For
now we’re stuck!
 Maybe malicious cold clumps are causing
the asymmetry, like the situation the
famous BN/KL high-mass star forming
region.
 Ultimately, we have to interpret the SED of
each HMPO.
Observed SEDs in IRAS
18089-1732

IRAS 18089-1732
100
IRTF - MIRSI
Spitzer - IRAC
B
10
A
Flux Density - Jy
IRAC and IRTF data suggest
moderately deeply embedded
source, not Av=1000 suggested by
mm dust continuum.
 Sources are on Gemini list for
spectra and astrometry.
 Ry started processing low
resolution spectra of HMPOs
between 8 and 13m.
 Cliff is beginning to make
spherically symmetric models to
get rough estimates of the sizes
and densities of the disks and
envelopes.
1
A
0.1
0.01
B
0.001
0
5
10
15
Wavelength - Microns
20
25
30
Models of deeply embedded low-mass protostars for
different angles of incidence suggest interpretations…
Whitney, et al. 2003 ApJ 598, 1079
What have we learned?



A significant fraction of
regions of high-mass star
formation have a single,
compact mid-ir source.
IR images of star
formation regions are
very hard to interpret.
SEDs are critically
dependent on the model
assumptions, e.g.
orientation of the outflow
cavity, and the shape and
density of the disk.
IRAS 19410 from GLIMPSE
What’s next?




Studying high-mass star
IRAS 18247-1147 from GLIMPSE
formation is like looking for a
serial killer.
We’ve got to look at each
crime scene (each HMPO’s
images and spectrum), looking
for commonalities and
idiosyncrasies.
We need to make computer
models that match the SED’s
and image sizes as a way of
estimating the physical
conditions in the disks, the
envelopes, and the outflows.
I hope!
In the end, patterns will
Next week, Dr. Lori Allen (CfA) will
emerge that pin down details
show more stunning Spitzer images
of the steps and their time
and some of the successes in the
scales.
study of low- mass star formation.