Evolution of circumstellar disks in premain sequence stars

1
Evolution of circumstellar disks in pre-main
sequence stars A proposal for the VO Science
Demo Enrique Solano (on behalf of the Spanish
Virtual Observatory)
2
General framework of the proposal

• VO has been identified as one of the priority
  lines for the Spanish funding body of
  Astrophysics. (Plan Nacional de Astronomía y
  Astrofísica)
• NOT Technology!!

3
General framework

• Project Development of the Spanish Virtual
  Observatory and its inclusion in IVOA.
• Aim Coordination of the activities of the
  Spanish institutes in the VO framework.
• Cornerstones
• Adaptation to the astronomical archives handled
  by LAEFF to the VO standards.
• Development of data mining tools.
• Education and Outreach activities in the VO
  context.

4
GTC
May2003
COROT
5
An example of data miningLight curve
characterization

• Visual inspection by experts is not feasible.
• Data mining techniques must be used instead.

• Network trained with HIPPARCOS light curves.
• Unsupervised topological mapping based on
  morphological proximity among the light curves.

More than 500000 objects in the input catalogue!!
6
(No Transcript)
7
Data miningLight curve characterization
8
The next AVO-Demo

• The AVO team needs more inputs from the SWG on
  the science cases. (P. Padovani, several times)
• Which way to go?
• Improvement of the AVO-Demo04 tool including new
  functionalities.
• Development of new tools covering other science
  cases.

9
The LAEFF proposal the science case

• Aim Characterization of the circumstellar disks
  of pre-main sequence stars by fitting theoretical
  models to the observed spectral energy
  distributions.
• ? Propelled by the discovery of extrasolar
  planetary system the understanding of the
  formation and evolution of protoplanetary disks
  is a crucial step.

10
The LAEFF proposal the science case

• Scenario
• ? Dusty disk surrounds the star in the first
  million years.
• ? Dust particles will start to aggregate to form
  planetesimals or planetary cores. The gas around
  them is swallowed and the protoplanets start to
  grow.

11
The LAEFF proposal the science case

• Open questions
• ? What is the process driving the evolution from
  dense disks in PMS to debris disk in Vega-type
  stars?.
• ? When is the starting time for the evolution?
• ? Does it always occur?
• ? What does it depend on?

12
Observational techniques

• Direct imaging both PMS and MS

13
Observational techniques (II)

• Spectral energy distribution widely used to
  study the properties of protoplanetary disks.
• Merín et al. (2004, AA, 419, 301)
• the proto-proposal paper
• ? Comparison the observed SEDs with a grid of
  theoretical models.

14
The irradiated accretion disk modelsDAlessio et
al. (2001, ApJ, 553, 321).

• Assumptions
• The disk is in a steady state (dM/dt constant)
• Geometrically thin (H/R ltlt 1).
• The dust and gas are well-mixed in the whole
  disk. Usual dust to gas mass ratio 1 / 100.
• LTE
• Abundances given in Pollach et al. (ApJ 1994,
  421, 615)
• Grain size distribution n(a) n0 a exp(-p)
  where
• ? 2.5 lt p lt 3.5
• ? Minimum grain size 0.005 µm
• ? Maximum grain size 1 µm to 10 cm

15
The library of models
Central Star Teff 4000 - 10000 K (K7 V - B9
V) Ages 1, 3, 10 Ma Disk dM/dt 10-6, 10-7,
10-8, 10-9 M?/year ? 0.01 Rdisco 800, 300,
100 AU Inclination angle i 30, 60º Dust n(a)
a-p , p 3.5, 2.5, amin 0.005 ?m amax 1, 10,
100 ?m, 1 mm, 1 cm, 10 cm
16
An example HD 34282

• disk radius Rmax 705AU
• mass accretion rate8.2e-9 solar masses/year
• inclination angle of 56

17
The physical interpretation

• Two grain sizes thin dust particles in the
  surface of the disk, and large particles in the
  mid-plane of the disk.
• The black body emission at 3µm comes from the
  inner edge of the disk, frontally illuminated by
  the star and not included in the disk models.

18
The aim of the proposal
Repeat this time consuming analysis in a much
more efficient and uniform way covering the full
EXPORT sample (very complete set for around 30
PMS stars).
19
The data resources The EXPORT consortium
(www.laeff.esa.es/EXPORT)

• EXPORT EXtrasolar Planet Observational Research
  Team
• Large international collaboration to study the
  formation and evolution of planetary systems.
• EXPORT was granted with all the International
  Time (5 of the total time) allocated in the
  Canary Islands Observatories in 1998.
• Simultaneous photometry from the optical to the
  near IR.
• Medium and High res. spectroscopy in the
  optical.

20
The data resources (II)
21
The system conceptual design
22
AVO-Demo04 What should be improved?System
Interface

• VO is something new
• ? The learning process requires some time.
• ? The system must be easy for the new users to
  learn BUT, at the same time, it must be able to
  cope with complex queries.
• The friendliness of the User Interface is
  extremely important for the system to be
  effectively used by the community at large.
• Use of the AVO-Demo04 is not straightforward.

23
AVO-Demo04 What should be improved?users
assistance

• On-line access to project documentation A
  detailed description of the project should be
  given on-line and stored in a format allowing an
  easy browsing.
• On-line help Help on a specific keyword can be
  obtained by simply clicking on it.
• HelpDesk To channel questions and to provide
  continuous support to users of the Archive.
• FAQs systems, workflows, on-line examples,

24
The system workflow

• Query via users interface.
• ? Object Id. Coordinates Archive Selection.

? Input magnitudes physical parameters.
25
Workflow input physical parameters

• Estimation of the stellar physical parameters
  (Teff, log g, M/H) and reddening.
• Pre-defined hierarchy. Values provided by the
  user on the top.

• Differences between the input and output
  parameters may indicate
• ? Non-adequate set of input values
• ? Presence of peculiarities in the SED that might
  deserve further and more detailed analysis.

26
Data mining tools

• The problem Given a set of points in a observed
  SED, how to obtain the physical parameters of
  both the star and the disk.
• The goal
• ? To determine the most a posteriori probable
  set of parameters.
• ? To determine what other alternative models can
  compete with it.
• ? To determine what are the most discriminant
  observations to discard alternatives.

• Fundamental added values to the system.
• Optimization of the Scientific potential.
• Key activity in the VO framework. This is from
  the greatest scientific benefits are expected to
  come.

27
Data mining tools (II)

• The method Model parameter estimation using
  Bayesian methods.
• ? Given a set of prior probability densities for
  the different parameters (inclination angle, disk
  radius, mass accretion rate, dust size
  distribution, maximum size for the disk, Teff,
  log g, M/H), to compute the posterior
  probabilities for the data.

28
Data mining tools (III)

• Implementation of an inference tool to
  quantitatively assess the best model degree to
  sufficiently explain the data.
• Occams razor philosophy a simpler theory with
  a more compact parameter space will have a larger
  evidence than a higher order (more complicated)
  model unless the latter performs significantly
  better at explaining the data.
• ? Star ? Star Disk model
• ? Star Disk Inner wall
• ? Star Disk-1 Inner wall Disk-2 (HD34282)

29
The LAEFF proposal summary

• General-purpose tool (not only disks)
• Support spectroscopy (started in AVO-Demo04).
• Access to local and remote multiwavelength and
  heterogeneous data (archives and catalogues).
• VO-compliant access using VO-standard
  procedures.
• Implementation of data mining tools.
• Access to theoretical models.
• Possibility of inclusion of user measurements.

30
The proposal Future applications

• The tool should aim to satisfy the needs of
  specialized users by
• ? Allowing inclusion of new data (Spitzer, MIRI)
  and new models (Vega-type stars).
• ? Allowing implementation of new tools.
• ? Close feedback with the users.
• ? Collaboration with individual scientist
  acting in universities and research centres as
  software tools writers.