ESIP Air Quality Jan 2006 1 Air Quality Cluster Technology Track Earth Science Information Part ners Partners • NASA • NOAA • EPA • (?) • USGS • DOE • NSF • Industry… Data Flow & Interoperability in DataFed Service-based AQ Analysis System R. B. Husar, S. R. Falke and K. Höijärvi Washington University, St. Louis, MO ESIP Federation Winter Meeting 2006 Washington, DC, January 4, 2006
16
Embed
0603 Esip Fed Wash Dc Tech Pres 060103 Esip Aq Tech Track
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ESIP Air Quality Jan 2006 1
Air Quality ClusterTechnology Track
Earth Science Information PartnersPartners
• NASA• NOAA• EPA
• (?)• USGS• DOE• NSF• Industry…
Data Flow & Interoperabilityin DataFed
Service-based AQ Analysis System
R. B. Husar, S. R. Falke and K. Höijärvi
Washington University, St. Louis, MO
ESIP Federation Winter Meeting 2006Washington, DC, January 4, 2006
Federated Network for Air Quality Data and Processing ServicesProject Team:
Software Architecture: R. HusarSoftware Implementation: K. Höijärvi
Data and Applications: S. Falke, R. Husar
The DataFed Project
ESIP Air Quality Jan 2006 3
DataFed in a Nutshell
DataFed VisionAid air quality analysis by effective use of relevant data
DataFed Goals
Facilitate access and flow of AQ data from provider to users Support user-driven data processing value chains
Participate in specific application projects
Approach: Mediation Between Users and Data ProvidersDataFed assumes spontaneous, autonomous data providersNon-intrusively wraps datasets for access by web servicesMediates, homogenizes data views. e.g. geo-spatial, time...
Applications
Browsers and analysis tools for distributed monitoring data Serve as data gateway for user programs; GIS, science tools
DataFed is focused on the mediation of air quality data
Programming Access, Process, Render Data by Service Chaining
[Station]
[Monitor]Clickable monitor locations layer
Map boundary layer
ESIP Air Quality Jan 2006 6
Single Data Model for All AQ Data
Most Views are slices through a cube of data organized by lat, lon, altitude, and time (X,Y,Z,T)
Multidimensional Data Cube
ESIP Air Quality Jan 2006 7
OGC Abstract Specification – Coverage (00-106)
• Abstract Spec.: Coverages model and visualize spatial relationships between, and the spatial distribution of, earth phenomena
• Coverage subtypes of Air Quality interest are Image, Grid and Discrete Point
• Has a property (Coverage_Function) that has a spatial domain, and a value set.
• For Image and Grid, the Coverage_Function is an array…Discrete Point coverage is described by a PointC_Function where the spatial domain consists of a collection of points
Observation:The Abstract Spec. foresaw the need for Point Coverage, i.e. fixed monitoring Stations
Earth Science – GIS InteroperabilityShared Data Model
ESIP Air Quality Jan 2006 9
OGC Web Coverage Service (WCS) Specification
• HTTP GET/POST based interfaces• Services have XML service descriptions (“Capabilities”, “Description”)• Filter parameters allow selection of subsets of source data• Output formats advertised by each service instance
OGC WCS getCoverage SchemaSuitable for wrapping with SOAP envelope, WSDL access, loose coupling
WCS is for "coverages" – information representing space-time-varying phenomena
WCS describes, requests and delivers coverages in spatio-temporal domain
WCS version 1.1 is limited to grids/"simple” coverages with homogeneous range sets
ESIP Air Quality Jan 2006 10
through
Data Access through Adapters
DataFed SOAP,HTTP Get
OGC WCS HTTP Get, Post
OGC WMS HTTP Get
Station-Point SQL Server,
Files…
Sequence Image, file
nDim Grid OpenDAP NetCDF,
…
Other Traject., Event, Pic
Sources Diverse formats
Many data models
Data Wrapper Data into geo-cubes
Queries to views
Virtual Data Cube Global geo-cube data model Makes queries data-neutral
Lat/Lon Box Elev Range Time RangeMap: BBOX=-180,-90,180,90, 1350,1350& TIME=2005-12-06/2005-12-06/PT3HTime: BBOX=-34,49.05,-34,49.05, 1350,1350& TIME=2005-12-05/2005-12-08/PT3HElev: BBOX=-34,49.05,-34,49.05, 0,18000 & TIME=2005-12-06/2005-12-06/PT3H
The form of the WCS query is the same for all slices through the data cube (views) The only difference in the views is the thickness of the slices in each dimension Return grid is in multiple formats (NetCDF, CSV, GML, PNG, … )
OGC WCS protocol is suitable for delivering most air quality data:
Suitable for most data view queries
BBOX, TIME
Station-Point SQL Server,
Files…
Sequence Image, file
nDim Grid OpenDAP NetCDF,
…
Applicable to most data types
Station, Image, Grid
For Air Quality data needs: • WCS: Add Point coverage to WCS (in addition to grid, polygon) • WCS: Strongly typed (XML Schema-verifiable) return data types• Testing: Much fixing and testing for server-client compatibility