Observing Simulated
Protostars with Outflows
arXiv:1205.0246
Stella Offner (NSF Fellow, CfA)
Thomas Robitaille (MPIA), Charles Hansen (formerly Berkeley)
Chris McKee (Berkeley), Richard Klein (LLNL/Berkeley)
Labyrinth of Star Formation, Crete 2012
Motivation
• How accurate are protostellar properties inferred
from SEDs?
• How does viewing angle, multiplicity, or stage effect
inferred properties?
SED
Models
Source
??
Enoch et al. 2009
age, inclination, star mass, disk mass, envelope mass, disk
“Best
Fit”
radius, outflow opening angle, accretion rate, density
profile, disk radius, envelope radius, stellar radius...
Outline
Observe
Model
Simulate
Compare
“Best Fit”
Methods
• Adaptive Mesh Refinement (ORION)
• Turbulence
• Gravity
• Radiative Feedback (model for stellar
evolution; Offner et al. 2009)
• Outflows use a model based upon Matzner &
McKee 2000 (e.g. Cunningham et al. 2011, Offner et al. 2011)
Simulate
Offner et al. 2012
Zooming
• 1 Base Calculation
• 1 freefall time
• 130 AU resolution
• 4 Zooming Calculations
• 4 AU resolution
• “0”, 15, 30, 60 kyr
Log Column Density
Offner et al. 2012
Gas Velocity
~1 km/s
~10 km/s
L=0.65 pc
Post-Processing
• 3D Dust Continuum Radiative Transfer Code
• Monte-Carlo
• Parallel
Robitaille 2011
• Inputs: dust model, density, sources (L, T)
Post-Processing
• 107 photons
• 21 Protostars at final time (D burning)
• 200 Wavelengths (0.01 µm - 5000 µm)
• 20 Apertures (1,000-20,000 AU)
• 5 Resolutions (4-65 AU)
• 20 Viewing Angles
Observe
Robitaille 2011
Resolution
1mm emission
65 AU
4 AU
Offner et al. 2012
Wavelength
Padgett et al. 99
Offner et al. 2012
SED Zoo
Offner et al. 2012
Luminosity (Lsun)
Bolometric Luminosity
Compare
]
20 viewing
angles
* Actual
+ Hyperion
Median
Multiples
Source
Offner et al. 2012
Bolometric Temperature
Class 0 = 0.1 Myr
Class I ~0.34 Myr
+ Different
Views
+ Median
Tbol (K)
(Enoch et al. 09, Evans et al. 09)
t = 30 kyr
Source
Offner et al. 2012
Model Comparison
Compare with Robitaille et al. 2006:
200,000 model library
■ “Observations” with
2MASS, MIPS, IRAC,
Bolocam
∇ Upper limits
__ Best fit model
__ Good fit models ( χ2< 3N)
Model
Offner et al. 2012
Model Comparison
Model/Actual
“Good Fits”
Symbols = “Best Fit”
Compare
Source
Offner et al. 2012
Model Comparison
Model/Actual
“Good Fits”
Symbols = “Best Fit”
Source
Offner et al. 2012
Model Comparison
Model/Actual
“Good Fits”
Symbols = “Best Fit”
Source
Offner et al. 2012
Model Comparison
Number of Models
Actual
All
Good Fit
10% Lbol,ρ
Offner et al. 2012
Model Comparison
Model/Actual
“Good Fits”
Symbols = “Best Fit”
Offner et al. 2012
Model Comparison
Model/Actual
Envelope
Mass
Expected based on
opacity differences
Symbols =
Different
Sources
Aperture (AU)
Offner et al. 2012
Conclusions
• SEDs and inferred properties are very sensitive to
the viewing angle
• Bolometric luminosity can range by x5
• Sources may span 2 classes even early on (ages overestimated in 5-10% of cases)
• Caution is necessary when extrapolating source
parameters from SED models, e.g.:
• “Good” accretion rates may span 2 orders of
magnitude, but usually centered close to actual
• Envelope mass may be x2 (or more) too high
mean luminosity from the models. The error bars on the model
lifetimes derive from the observal luminosity uncertainty. The two
observational results with uncertainty are shown by the thick set
of solid error bars.
Shameless Plug
The Protostellar Luminosity Function
Evans ea 09
Isothermal Sphere
Two-Component Turbulent Core
Turbulent Core
Competitive Accretion
T
N
lu
tu
lu
L (Lsun)
n
ca
Offner & McKee 2011 p