TSAP: an SVO-ESAVO collaboration to develop a protocol to access theoretical

data.

Application to PGos3 Enrique Solano (LAEFF/SVO)

Interop., Kyoto, May 2005

SVO Carlos Rodrigo Miguel Cerviño

ESAVO Jesús Salgado Isa Barbarisi Pedro Osuna

Interop., Kyoto, May 2005

Final goal: To ensure a real interoperability between

observational and theoretical data.

Basic requirement: Specific VO-theory standards must be

defined. Some already existing VO standards are useful (VOTables,

UCDs, data models,..) but some others are not (e.g. access

protocols).

Theory in the VO

Definition of a new protocol for theoretical data

“In the light of the positive experience of the approach in the

IVOA it will likely be very useful to define some simple query

protocols for simulations as well and this effort should be one

of the first for the TVO to undertake” .

(Lemson & Colberg, Theory in the VO white paper)

Interop., Kyoto, May 2005

The problem:

Observational data can be unambiguously defined by their

coordinates. They are, therefore, prime parameters for SIAP,

SSAP AND Conesearch.

For most of the theoretical data, Cone Search, SSA, SIAP

cannot be applied as absolute positions on the sky are

irrelevant. Moreover, the input parameters are model-

dependent: there is not a common set that can play the role

of the coordinates for the observational data.

Theory in the VO

Interop., Kyoto, May 2005

TSAP sounds familiar (very much like SIAP or SSAP) and it is simple and easy to implement.

TSAP: an access protocol for theoretical data

The answer: the “three-steps” process:

The server accepts the “FORMAT=METADATA” request giving back the list of input parameters (and appropriate descriptions) in VOTable.

The client shows them to the user who makes the selection (fixed values or ranges of values).

A new SSAP-like query is submitted and the results are obtained.

Interop., Kyoto, May 2005

Data model for theoretical spectra

To ensure an efficient comparison, it is advisable

that any data structure proposed in the

theoretical domain resembles that proposed for

observational data.

The output VOTable is as close as possible to the

IVOA SED Data Model v0.93. Although Coverage

and Frame are meaningless fields…

Spectral coordinates and flux values. Curation fields

DataID fields

Interop., Kyoto, May 2005

The server:

A practical case The client:

Interop., Kyoto, May 2005

Name resolved by SIMBAD

Registry Access (SSA) +

Theoretical services

Selection of input parameters Description attached

to each parameter

Interop., Kyoto, May 2005

Comparison of observational and theoretical spectra using

VOSpec

Model data scaling

Interop., Kyoto, May 2005

What’s next?Application I: accretion disk models

Interop., Kyoto, May 2005

Application II: PGos3

Federative project that plans to include public codes

related to the modelling of stellar populations in

galaxies.

Beta version. The development and implementation

phase began in July 2004 during the Guillermo Haro

Workshop: “Violent Star Formation and the Legacy

Tool” held at INAOE, Mexico.

Synergy between astrophysicist ands computer

scientist in Mexico and Europe.

http://ov.inaoep.mx

Interop., Kyoto, May 2005

PGos3 in the VO framework

PGos3 is a very valuable tool in the VO context:

'Star Formation Histories in Galaxies' was the extragalactic

Science Case chosen for the AVO-Demo 2005. This involves

retrieving multiwavelength datasets that are then compared

against stellar spectral evolution codes, to enable information

to be derived on galaxy star formation histories.

The Spanish Virtual Observatory is adapting PGos3

to the VO standards and requirements.

Interop., Kyoto, May 2005

Definition: a model of the expected emission of a

stellar population defined as the sum of the emission

of the individual stars.

The evolutionary synthesis models

The building blocks:

Initial Mass FunctionInitial Mass Function (IMF): the

number of stars with different masses

to include in the population at t=0.

The IMF is often approximated either

by a power-law or by a sum of power-

laws over different subranges.

(e.g. α: 2.35, Salpeter, (1955):

Interop., Kyoto, May 2005

The IMF must be translated into spectral information to obtain the population spectrum

at t=0. Stellar Model AtmospheresStellar Model Atmospheres

Evolutionary synthesis models (II)

Interop., Kyoto, May 2005

TSAP in PGos3: an example

Interop., Kyoto, May 2005

Next step: implementation of the protocol for evolutionary synthesis

models