Transcript UK AstroGrid and STP Chris Perry, Rutherford

Solar Terrestrial Physics:
ESA SpaceGRID
Chris Perry, Rutherford Appleton Laboratory
21st March 2003
STP: SpaceGRID
• Introduction
• Existing Activities
– Community Consultation
– AstroGrid: Phase A - STP Pilot
– ESA SpaceGRID - Solar System Research
– Related activities
• e-Science and other opportunities
Introduction
• Solar System Research is a multi-disciplinary science
– Solar Physics, Solar Terrestrial Physics, Planetary
• Complex 3-D environment
– Phenomena occur over a range of temporal and spatial scales
– In general observations are not object based.
• Complex set of instrumentation, data and formats
– Particles, fields, waves and imagers
– Scalers, vectors, tensors, images, multi-dimensional arrays
• Data processing is often responsibility of PI
– Heterogeneous data handling systems
• Researchers need to combine and manipulate multiple data sets
Community Consultation
• Community consultation via online questionnaire
• Results and discussions held at MIST, EGS and at project meetings
• 113 responses from 20 countries (38 from the UK)
• 55% Solar, 36% STP and 9% Planetary
• Broad spread of interests including:
– planetary interiors & surfaces, planetary and terrestrial atmospheres,
ionosphere, aurora, magnetosphere, solar wind, solar terrestrial,
interaction, solar flares, corona, solar atmosphere and helioseismology.
• Most people who responded are working on data analysis though
significant minorities involved in modelling, operations and software
development.
Community Consultation
• A few of the difficulties highlighted in data access:- Locating sources of data or models
- Ascertaining data availability for multiple data sets
- Limited search capabilities
- Multiple requests to different archives, interfaces, access control
• Main problems combining data from instrumentation and models:- Lack of standardisation in data specification.
- Inadequate, incompatible or unreliable metadata.
- Need for re-sampling temporal or spatial data.
- Lack of documentation -> incorrect use of available data.
Community Consultation
• User prioritisation of baseline requirements
All SSR
Resource Location
Metadata
Data
Compatibility
Catalogue Query
Data Query
Dist. Cat. Query
Dist Data Query
Remote Processing
Web Portal
GRID Server
API
Collaborative Env.
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
AstroGrid Phase A: STP Pilot
• Simple test-bed to help assess requirements
• About 10 data sets from each of CDHF for STP (ACE
and GOES KP) and WDC (geomagnetic indices, Dst,
aa).
• Hybrid solution using a combination of existing CGI
based software and newer SOAP based technology.
• Not web/grid service based (I.e. no WSDL).
• STPDF toolkit used to provide uniform view of
heterogeneous data sets.
AStroGrid Phase A: STP Pilot
•Allows selection of system or user
datasets.
•I/F dynamically updated to show
available fields and number of
records for selected data set.
•User datasets are the result of
previous operations allowing
compound queries to be built.
Four basic operations supported
1) Select/Output fields
2) Query data set
3) Find/Select time intervals
4) Time series join (nearest
neighbour).
AstroGrid Phase A: STP Pilot
Simple UI to select parameters to be plotted and the
time range. In this case a joined data set consisting
of ACE MFI from CDC and Dst from WDC (one
years worth of data).
Clicking on resulting plot, re-plots with a shorter
time range.
Simple ASCII output of selected range that can
easily be loaded into a spreadsheet.
ESA SpaceGRID Study
• SpaceGRID is an ESA study aimed at identifying the
potential benefits of Grids" to the ESA community and
defining a road map for the implementation of this
technology within ESA.
• The SpaceGRID study spans a broad range of
domains.
–
–
–
–
Earth Observation
Various aspects of spacecraft engineering
Space Weather
Solar System research (solar, STP and planetary).
ESA SpaceGRID Study
Survey of SSR
Facilities
SSR Baseline
Requirements
Evaluate
Options
Capture User
Requirements
Started Oct 2001
PDT 20th March 2003
Final presentation May 2003
Dissemination
Activities
Plan Prototype
Implementation
Prototyping
(led by SSL)
ESA SpaceGRID: SSR
Prototype
• Implemented by SciSys and RAL
• Grid Services using OGSA (GT3 Alpha 2)
• Standard XML VOTable used to hold query
specification and returned results
• XML based metadata catalogue, using XSLT to
search/transform to VOTable result.
• Federated several Solar event catalogues, Cluster
Summary Parameter data and World Data Centre
parameters.
ESA SpaceGRID: SSR
Prototype
• Prototype Demonstration Test held at ESTEC
yesterday
• Three data archive hubs (two at RAL and one at
ESTEC)
• Each hub provides a set of simple services:
–
–
–
–
Catalogue Service
Data Mining Service
Query Service
Storage Service
• The Query Service interacts with legacy archives
using a simple file based interface (data centres don’t
need to learn OGSA).
Related Activities
• SPASE - Space Physcis Archive Search and
Exchange
– Working group involving international Space
Physics data centres (GSFC, JPL, APL, SWRI in
the US, CDPP and RAL in Europe)
– Developing standards (e.g. a metadata dictionary)
for the Space Physics Research domain.
– Standards are provide the basis for system
interoperability and are critical to the success of escience/Grids.
PPARC e-Science
•
•
•
STP Data Centres
– World Data Centre (UKSSDC) - Solar System Data Warehouse
– UK Double Star Data Centre
UK SPASE/ UK SPASE Portal
– Formal/funded participation in the development of international
Space Physics interoperability standard.
– Portal into national and international Space Physics Data Centres
(e.g.. UKSSDC (WDC), ESA Cluster Archive, ESA Planetary
Archive, CDPP, US SPASE compatible archives)
UKEEDC
– A UK Data centre for the ESA Eddington mission. (within the
AstroGrid consortium? )
Other related e-Science
activities
• ESA Cluster Active Archive
– ESA SPC approval on 25th Feb 2003
– Interoperability (in particular with ILWS) included in initial
design and may be Grid based.
• Possible follow-up to ESA SpaceGRID to be
announced in Autumn
• EU Framework 6