YSO/PMS disk types, time-YSO/PMS disk types, time-scales and evolutionscales and evolution

Luke Thomas Maud – ESAC Trainee 2009 Bruno Merin – Herschel Science centre

(Supervisor) Hervé Bouy – Herschel Science centre

(Research Fellow)

OutlineOutline Formation Scenario The Sample Spectra Categorization Mass and Age Estimates Results Conclusion and Possible Interpretation The Future

YSO FormationYSO Formation Young stellar objects

form in molecular clouds, opaque to optical light

Clouds collapse, cores form..

Angular momentum conservation generates a spinning disk as material falls inwards..

As the disks evolve they disperse/evaporate and the central star joins the main sequence...

It is possible that a planetary system is created.

Using IR we can probe clouds and ‘see’ YSOs inside:Using IR we can probe clouds and ‘see’ YSOs inside:

Formation – Spectral Energy Formation – Spectral Energy DistributionsDistributions

Since the late 80s investigations have been undertaken to evaluate the evolution of YSOs

Adams, Lada and Shu 1987 developed the ‘classical’ classification system for YSOs

Using Spectral Energy Distributions (SEDs) one can see how the flux of the emission varies

Thus how the YSO evolves This however is the ‘general’ model and

classes are based upon a slope from 2 to 24 um wavelengths

The SampleThe Sample

This project is a continuation of previous papers investigating the 5 close star forming regions from the Cores to Disks (c2d) Spitzer legacy program (Evans et al 2009)

The initial sample had photometry for 1024 YSO candidates with estimated ages of 1 -10 Myrs

Encompassing the timescales of disk dissipation previously observed for low mass stars and probing slightly different environments

IC 348 Perseus

Rho Ophiuchus

NGC 1333 Perseus

Serpens Core

The SampleThe Sample However after analysis, the C2D YSOs had

unexplainable age attributes The problem is due to the degeneracy involved

in SED fitting We created a new sample of objects that had

full spectroscopic data, this means the Spectral types of the stars could be constrained

We include only stars that are Class II and III from (Greene et al. 1994)

II -1.6 ≤ α < -0.3, III α < -1.6 We now have a fully usable sample of 819 YSOs

Spectra Energy Distributions – Spectra Energy Distributions – (Spectra)(Spectra)

For all the targets in the sample we create SEDs Thus we plot the wavelength of photometric

data vs. the flux at that wavelength

ClassificationClassification In comparison to the basic Class II and III we

see a more diverse range of SEDs:

ClassificationClassification So we define limits of classification with reference to

the median CTTS and the fitted photospheres

Mass and Age EstimatesMass and Age Estimates We get temperatures from the spectral

types and luminosities from the fitted photospheres

These are passed to HR diagrams for plotting against PMS tracks computed be Baraffe et al (1998) and Siess et al (2000)

Sanity check that we are not doing anything ‘silly’ as the objects fit on the HR diagram.

Mass and Age EstimatesMass and Age Estimates We group the objects into 4 groups now

Under Luminous – Sources below the tracks – Need to be further away

Over-Luminous – Sources above the tracks – Need to be closer

Strange – Objects with non-fitting and ‘weird’ SEDs – Would require extra analysis

Final – Objects used in the main results – 664

• These 664 targets have excellent SED fits and have mass and age constraints from the HR fitting:

• Resulting SEDs: Good (81%)

• Over-Lum (5%) Under-Lum (5%) Strange (9%)

ResultsResults The resultant mass and age spread is

sensible, and within expected values for Low mass star forming regions:

We appear to represent stellar masses down to ~0.03Mo and up to ~ 3.5 Mo

ResultsResults We divide the final group in Mass ranges of:

M < 0.5 Mo, 0.5 < M < 1.5 Mo and M > 1.5Mo Separating the stars’ physical properties

Fully convective, radiative core/convective envelope and convective core/radiative envelope respectively

And Age ranges of: 1 – 5 Myrs, 5 – 10 Myrs and 10 – 20 Myrs

To cover the general disk dispersal time scales and evolution

We produce 9 pie charts rich with information

Conclusions - Interpretation

The Spitzer data allows the new categorization of the evolutionary stages of disks around YSOs via SEDs and clearly detects ‘transitionals’ Largest statistical sample 3-5x bigger than any previous and

covers a larger range of ages and environments

Initial stages all appear the same (fisher test) – NEW RESULT!

More massive stars remove there disks faster, none are present after 10 Myrs

Primordial class disks are still evident at ages above 10 Myrs for stars with M < 1.5Mo

Mass plays a role in evolution – Has been suggested but this is evidence!

Speculation

As it appears that disks around lower mass stars last longer more time for instabilities and coagulation of particles potential home for rocky planets?

The most massive stars’ disks disperse fast timescales preferential for large ‘Jupiter’ planets that

form over short times in rich disks (we need disk mass)

Latest statistics (Udry et al. 2007) suggest low mass stars have low mass multiple planet systems while high mass show single migrated Jupiter mass ones (obviously still limited in technology)

Future The age constrains are the

current best with simple HR fitting, however do not account for accretion luminosity and a more robust method is required (under-luminous if not

accreting = older)

To calculate disk mass we require observations extending into the sub-mm Herschel will provide this

coverage and add a new perspective allowing comparisons of disk and stellar mass. One of the key parameters of the disk itself

Our method is applicable to the Gould belt clouds currently underway With a much large sample

High resolution spectroscopy would help identify transitional disks Binaries are ‘empty’ vs.

planet harbouring disks

Alma and JWST will shed extra light on the data with increased sensitivity/resolution Imaging will allow disk radii

to be constrained

Thank You for your attention

Any Questions Please

A Problem? - Scaling Note the ranges probed by the photometry

(MIPS1) F/A type at ~ 21AU Solar type at ~ 7 AU Low-mass at ~ 3 AU

The previous assumptions assume the disks scale with mass and are all comparable – however they may not scale directly (1:1)

Thus for are low mass sources in the pie charts classified using a different range in the SEDs??

Scaling

NO, due to the nature of SED with the logarithmic scale we see only a small shift of data points

This would therefore supports that stars and disk may scale together as one may expect (more mass, more gravity, more disk)

Additional Classification

To move allow another perspective independent of the larger fluctuation of age in HR fitting; we also present a model based upon the scheme of Cieza et al (2007)

Cieza use turnoff and α excess – which describe the distance of the inner disk to the central star and the degree of opacity or thickness of the disk

These are grouped for the same masses as the pie charts

Additional Results The ranges of the 3 mass groups appear to be comparable

Suggests the actual evolutionary sequence maybe the similar independent upon mass

Note the tight range of Primordial disk, with Transitional disks above α = 0 and Settled moving right and down

Again the aforementioned ‘scaling’ problem is dismissed as one would expect the turn-off ranges to be altered, but clearly they are similar