|
Personal notes on the |
General presentation and concerned of the different WG (Working Group) and IG (Interest Group)
(the link of each WG or IG leads to their twiki web pages)
(Jonathan McDowell)
- STC standard
- Model for representing Atomic & molecular line (first draft)
- Catalogues & source catalogues (reconcile the different approaches)
- Characterisation Model (where is your data : where in space and time ...)
(Guy Rixon)
- Security issues & progress
- VOStore progress (reconcile VOSTore and VOSPace, issues and plans)
Applications (IG)
(Marc Allen -[CDS-strasbourg])
- (new and upcoming) VO tools services & libraries
=> interoperability of the different tools, how can they work together ?
- Capture ideas about how astronomers will use VO applications
- What information/metadata would be useful for doing so ?
(Andrea Preite Martinez -[CDS])
UCD : Unified Content Descriptors
- Deals with the description of quantities & meaning of what you can find in the Tables (originally, now its role actually expended from this)
=> maintain a list of [UCD] words :
- Semantic : how to go toward an IVOA standard procedure & vocabulary ?
- Explore Ontologies : first tests
(Roy Williams)
- Time critical (astronomical events such as TOO) i.e., "I have found <something> [that is, a new event] in the sky : how to observe it and follow it up afterwards ? "
(Francois Ochsenbein)
- News about few applications : TOPCAT, VOFilter + VOTable & HTML (VOTable2XHTML)
- How to describe COOSYS ? SAVOT (Java VOTable Paser)
COOSYS : COOrdinate SYStem Taken from the VOTable Format Definition V1.1 here is an example to show how this xml element looks like as used in the VOTable standard:
<COOSYS ID="J2000" equinox="J2000." epoch="J2000." system="eq_FK5"/>
(Tony Linde)
- Coming up with the metadata to describe the applications
- Allow to locate, get details of, and make use of, any resource located anywhere in the VO
Theory (IG)
(Gerard Lemson)
- produce Use Cases with matching to WGs
- extract list of theory specific requirements
Data Access Layer - DAL
(Doug Toby - [NRAO])
- Simple Spectral Access [Protocol] : SSA[P]
- Simple-Spectral Line Access [Protocol] : SLA[P]
- Simple-Image Access [Protocol] : SIA[P]
(Maria Nieto-Santisteban)
- review ADQl & SkyNode Specifications
Data Models - Jonathan McDowell
=> STC & STC-lite (subset of the STC) - Arnold Rots
There was some controversy whether the existence of both STC versions is well-founded. Indeed these two versions are incompatible though having similar schemas. This creates some confusion
=> IVOA [Source] Catalogues - Pedro Osuna
Where source: any object that radiates above the ambient radiation (i.e., it could a nebulae, a star, a galaxy ...)
Later on, we can imagine that in Source we have an attribute called <-type> that would point to other (type of) catalogues depending on its value :
Ex: Type = galaxy => Galaxy Catalogue (or Data Model)
PPN => PPN Catalogue (or Data Model)
...
just like the attribute <-coordinate> in <Source> point toward the STC & Observation Data Models i.e. :
=> Characterization Model - Mireille Louys
It is basically the Data Model that characterizes ONE Observation. The 3 "axis" characterisation (therefore called characterisationAxis) are respectively:
Bigger ‘picture’ of the Characterisation Data Model :
There are 4 levels of complexity/detail for each axis, for which the last level (level 4) describes the variability along the concerned axis => Level 1-2-3 : Meta discovery
Level 4 : more "subtle" details
(cf. here for an UML diagram figure and a general description of the Characterisation Data Model)
/!\ at the moment the dependencies along the different characterisation axis have not been developed
In order to account for the errors on each axis (and refine them therefore) another ‘bit’ has been added - the AxisFrame :
New features : - errors on the data (the typical error)
- sampling (samplingRefVal class) a level 1
/!\ Observation here is a reduced dataset
Characterization XML schema (Sept 2005 version) can be found here
=> Line Data Model - Matteo Guainazzi
This model defines the standard to access any atomic & molecular line databases
Spectral Data Model (SDM)
Applications IG - Mark Allen
=> New functionalities in VOSPec - Isa Barbarisi (ESAC)
The idea behind this application is to unify the 2 following protocols :
Theoritical Spectral (TSAP) & Simple Line Access Protocol (SLAP)
so that one can compare more easily theoretical spectra with observed ones.
TSAP : cf. ISO Astronomical Spectroscopic Database (IASD)
Cf. here to access VOSpec itself and have a real-time demo (via a JAVA enable brower).
=> VOSed - Raúl Gutiérrez (from SVO : Spanish VO)
This is a tool for building Spectral Energy Distributions. It also allows comparison with theoretical models
Cf. here for a taste of this application.
SOAP: Simple Object Access Protocol (Protocol used for communication in Web services)
=> VisIVO - Claudio Gheller (INECA)
This application is for visualization of N-dim (N>2) VO data
This is an Open source, MS window (porting to Linux 1.0) application.
At the present time it interact/interoperate with Vizier and Aladin dynamically
Cf. here for a full description of that tool and to download it !
=> Experience gained by enabling Interoperability in Aladin - Pierre Fernique (CDS)
Aladin & JAVA development
=> Python Tools for VO : PyWS & PyVOTable library - Franscesco Pierfederici
PyWS is a Python SOAP toolkit for inspecting and "consuming" VO Web Services. It creates a local proxy to the remote application
PyVOTable is a high performance Python VOTable Library. It is available here.
WSDL :Web Services Description Language
=> Workbench & ACR - Noel Winstanley (AstroGrid-JBO)
The workbench is a [JAVA] Web Start application. It is a suite of GUI built on ACR. Here it is.
ACR : Astro Client Runtime is a desktop service that makes it simple for other programs to access VO services.
=> VO - enabling "xmdv"- lite - John Taylor (INECA)
xmdv is a tool for viewing multi-dimentional datasets. It is NOT an astro-specific functionality.
The idea behind this tool is to use JAVA Web Start to use the library of other applications.
As for the other application : see here for a trial !
Theory [interest Group] Use Cases - Gerard Lemson
=> SVO -ESAVO use Case - Enrique Solano ()
- Definition of Science Use Case to extract data from the various VO Web Services.
- Make use of the TSAP (Theory Spectra Access Protocol).
Example of evolutionary synthesis model tool : PGos3 (Mexican tool). The idea behind this Use case : One could (using VOSPec for instance) compare between the results given by different grids of evolutionary synthesis models (for instance the PGos3 + other similar simulation tool) with the observational data.
=> VO Theory Use Cases - intermediate scale - D. De Young (US-NVO)
- N Body simulation of Globular Cluster Evolution
- MHD simulation of collimated outflow from AGN
=> Tool for Cosmological simulations Use Case - Laurie Shaw ()
Tool for cosmological simulation so that the output is an image in X-ray ....
=> Determination of physical conditions in interstellar clouds & Multi wavelength analysis of interstellar clouds - Franck Le Petit (Observatoire de Meudon)
- 2 use cases concerning the IMS
=> Theory SkyNode - Gerard Lemson ()
DAL - Doug Toby
DAL : Data Access Layer
Dealing with catalogues, images, spectra, SEDs, time series, events & variability data. It is about their access via the VO
- Access is analysis oriented
static (archive) vs virtual (subset/transformed data)
Data Model based access
DAL is object-oriented =>
- Object hierarchy
=> characterisation & curation of the generic dataset
=> SSA[P]- Doug Toby ()
SSA[P] : Simple Spectral Access [Protocol]
- 1D spectral, time series & SEDs
- Dataset creation type :
- Atlas <=> uniform survey
- pointed <=> variable instrumental data
- cutout <=> data sample not modified
- resampled <=> data sample computed by service
here it was a disagreement about 'how' the dataset should be qualified if it is coming from a survey (i.e., the term <pointed> particularly but also the term <Atlas> seem to confuse people ..)
- *Dataset derivation :
- Observed
- Composite
- Dynamic <=> simulated observation from real data
- Synthetic <=> simulated obs. from theoretical model
* : i.e., gives information about the type of observations dealt with
here-below is the general structure of the SSA[P] Data Model used in USA looks like :
Data Model used in USA for
Spectral data, dataset, Target, Curation, Characterisation, provenance (instrument it comes from)
Mandatory query parameters (for the query interface [i.e., mandatory for the SERVER not for the CLIENT])
- POS <=> RA, Dec
- SIZE <=> diameter (decimal degrees)
- TIME <=> date1, date2 (decimal years UTC)
- BAND <=> wave1, wave2 (m in vacuum; source or observer)
- FORMAT <=> VOTable, FITS, XML, text, graphics, HTML, external
Recommended query parameters
- APERTURE[1] <=> approx. spatial resolution
- SPECTRES
- TOP <=> number if top-ranked records to return
- OBJTYPE <=> mandatory if the service can return multiple object types
- COLLECTION <=> data collection identifier
[1] : general agreement that this parameter name should be changed - since it is fairly misleading - by something like SPECT_RES
[2] : comment from ?? : a time range parameter is missing (i.e. "give me all the set of observation of my favourite source that span 2 years (i.e., no matter what is the starting or finishing date)")
optional parameters
Comment : - missing selection per name. TargetClass (i.e., that allows you to choose a Target type such as <galaxy>, <pulsar>, <PPN>, <QSO>, <AGN> ..) nut no parameter for finding a specific object. This raise the problem of a Name resolution.
Query response
- Query Metadata (metadata describing the query itself), Target Metadata (astronomical target) ...
- Characterization1 - coverage, Characterisation2 - Accuracy, Characterisation3 - Reference Frame
Data retrieval
- Data Format : VOTable, FITS, XML, text, graphics, HTML, external
=> Spectrum Data Model [SDM]- Jonathan McDowell(CfA US-NVO)
SDM : Spectrum Data Model
That is Data Model that lies behind the query
ACD: Automatic Call Distributor (is a telephone facility that manages incoming calls and handles them based on the number called and an associated database of handling instructions)
WCS: World Coordinate System - added support to the WCS is needed (explicit listing of coordinates has been required)
STC - improvement is needed to match the current characterization & STC models
=> Characterization utypes "à la" XPATH - Francois Bonnarel(CDS)
utypes allow the software to deal with your data, i.e., gives a correspondence in the Data Model.
SSA input parameters that has something to do with the Characterisation (Data Model):
- 4 out of the 5 the required parameters (i.e., POS, SIZE, TIME, BAND)
- 2 out of the 5 recommended SSA parameters (i.e., APERTURE, SPECTRE)
comments : - may be a parameter characterizing the flux is missing at the moment, same for sampling (i.e., to characterise whether it is uniform or not).
Reminder: Resolution & samplingPrecision ARE NOT PART of Accuracy in the Characterisation Model
=> SLA[P] - Jesus Salgado (ESA/VO)
SLA[P] : Simple Line Access [Protocol]
3 activities :
- Line Data Model
- SIAP
- Test implementation (both server & client services)
from those 3 activities 2 clear divisions have to be made :
- Observational spectral line databases : lines observed and identfied in real spectra collected by different instruments/projects
- Theoretical spectral line databases
=> SIA[P] - Doug Toby
SIA[P] : Simple Image Access [Protocol]
it is based on the SSA work,
- expanded Query interface
- more Rich Query Response (generic dataset identification & characterisation + metadata extension qualification)
- selected feature (e.g., query response can be ordered by "score")
Future development
- Service verification
- Grid capability (i.e., data "staging", data management, web service interface)
=> SIA issues - Francois Bonnarel
- Availability problem solved by SIA V.1.0 : there is now about a dozen of services in the VO registries (such as SDSS, MAST ...)
- Also SIA V1.0 has an increased graphical capability now => toward SIA V1.1 which should contain in particular basic selection of images. The new SIA version will include dataset, curation, target, characterisation (level 1& 2 out of the 4 levels that the characterisation structure include), access description like in SSA.
- It is intended to cover also the specific features/case of the WCS concerning the basic usage of the graphic formats (i.e., "simple" WCS)
WCS : support for graphic formats in SIA
- Some extensions of the SIA have to be thought of (i.e., they should not be in the main SIA section/characterisation but still they should be (allowed to be) present in Extension Use Cases. Indeed Extension mechanism is needed for describing complex or heterogeneous metadata.
Principles : Main resources like SIA V1.0 or SSA V0.9
Additional resources => admitted including new tables, groups & XML elements
and then, using utypes you qualify the associations to the main section
on going development : implementation of the capability to have link to spectra of the selected region of your selected images (i.e., linking to services delivering spectra)
Plenary - Jonathan McDowell
The main problem : STC - STC-lite & Xinclude
=> Grid & Web Services - Guy Rixon ()
VOStore independent of VOSpace
DIME won't be the mendatory data transport (i.e., it will not be the standard, though allowed)
VO Basic profile : Covers SOAP
Security : will accept both VO community sites & Grid communities credentials (called "Certificate Authority" CA)
will accept "MyProxy" as a required interface on User Authentification Services (UASes)
=> Applications - Mark Allen (CDS)
7 applications were presented from actual prototypes to already 'usable' ones
standards used : SSAP, SLAP
call to registries to get available services
VOTable
Scientific Functionalities : may be more feedback concerning the scientific usage would be helpful
Interoperability between the different VO applications : ex. Aladin-VOPlot-VisIVO
A discussion was also held concerning the general mechanism : ex. "PLASTIC" has been proposed for providing interoperability between the various applications. Some more discussion on other mechanisms should be held.
=> UCD - Sebastien Derriere ()
Futur : maintain the list of words
Deal with Semantic : IVOA standard vocabulary
Explore Ontology : though Use cases
=> VO Event - Roy Williams ()
"Nothing special to declare"
=> VOTable - Francois Ochsenbein
The use of utype attribute (XPATH analogy) to reference to a Data Model is been brought but there is not a consensus on it (a substantial part of the audience found it "heavy")
=> Registry - Tony Linde?
- Description of applications via metadata has been discussed
- Resource metadata V1.1 should move on PP (Proposed Process)
- Demo of ESA VO Registry (by Aurelien Stebe) including validation tools
led to a discussion about
- curation
=> Theory - Gerarg Lemson
=> DAL - ?? ??
=> VOQL [& SkyNodes] - (Yuji Shirasaki &) Maria Nieto-Santisteban
ADQL + CORE + Extension
Core : provides easy integration of DAL & VOQL
Basic SkyNode will support core ADQL + Extension
Q: Specification of the table names qualified by archive name : how to do it ?
metadata need to be well defined
VOQL needs diversity
=> Final synthesis - Masatoshi Ohishi
the UCD Working Group (WG) is been actually renamed to Semantics WG !
End of meeting !
Cf here for other VO linked web pages and documents