Summer 2007 Session: Applying Service-Oriented Architecture Concepts to USGEO Near-Term Opportunities
The USGEO Architecture and Data Management Working Group (ADM) proposes to hold a session on applying service-oriented architecture concepts to USGEO near-term opportunities. The ADM is currently preparing a white paper on the benefits of a service-oriented architecture approach to the Integrated Earth Observation System (IEOS), the US contribution to the Global Earth Observation System of Systems (GEOSS). A basic tenet of GEOSS, and of IEOS, is that interoperability is required for success, and that interoperability pre-supposes integration at least at the level of the interfaces between components.
There are two observing system concepts (referred to as Near-Term Opportunities, or NTOs) which are far enough along to help the ADM with a proof of concept project. They are the Air Quality Assessment and Forecasting (AQ) NTO and the National Integrated Drought Information System (NIDIS) NTO groups. Toward that end, the ADM is organizing a two-part session during the open meeting day. The first set, to be held in the morning, will consist of simultaneous, but separate, meetings to answer the following question: “If I ruled the world, what would I in [AQ/NIDIS] need from you [NIDIS/AQ] that would make my job easier, and what could I contribute to you that I think would help you?” Put another way, what cross-community applications do not currently exist, but should. For instance are there models and model outputs that can benefit both communities, such as coordinated climate and meteorological models? Are their shared needs for certain types of data such as land cover data? Are there known interactions between the domains, such as drought-related impacts on air quality? Are there shared interests in and overlapping needs for services such as geospatial display and analysis of information?
Results from the morning sessions will be reported out at the afternoon joint session, at which the consensus opinion of the group will vote on the ‘services or service mashups or application(s) most likely to succeed.’ Time permitting, we will then try to do some designing, and possibly building, of those pieces we think are most viable. Ultimately, we’d like to be able to take advantage of the collaboration potential in the room, if not actually produce a beta-level tool that the Federation could then prototype and test.
Participants should read the IEOS Strategic Plan, the Air Quality and NIDIS Near-Term Opportunity Documents, the ADM Roadshow presentation, and any of a variety of information pertaining to service oriented architectures prior to attending. Most documentation is available at http://usgeo.gov; follow the links for Documentation (top left) or for the Architecture and Data Management Working Group (at the bottom left).
Who should attend?
- applications folks - experience with design of web-based applications
- discipline folks - domain experts on the science side who understand where the gaps are in the available data, models, and decision support tools - if we have some cross-discipline folks or multi-discipline folks, cool!
- web services people - familiar enough with the service side to talk about what's realistic
- SOA mavens, mashup builders, architects...
***New*** Draft Agenda
Summer 2007 ESIP Meeting: Applying Service-Oriented Architecture Concepts to USGEO Near Opportunities. July 19th
Morning session 8:00 12:00 Afternoon session 1:30 5:30/6:00
Breaks as needed throughout the day – nothing scheduled
Morning Sessions: AQ and Drought (NIDIS) held simultaneously in separate rooms
8:00 – Welcome, Introductions, Purpose, and Charge to the Group
Answer this question: “If I ruled the world, what would I in the (Air Quality or Drought Community) need from you (Air Quality or Drought Community) that would make my job easier and what could I contribute to you that I think would help you? “ And brainstorm on possible ways to make that wish a reality [i.e. build the tool you need by the end of today].
8:20 – 11:00 – Small Group Sessions – Use these questions to help answer the above.
1. What data, models, applications already exist that can be of greatest benefit to the Air/ Drought) Community?( objective is to get info on List known models, analytical tools, visualization/display tools, and data resources needed for the domains
2. What key questions does this community have to answer?
3. What are the gaps? What needs to be developed?
4. How will the nations citizens benefit from these existing and possible future capabilities?
1. What are the several (1-5 or so) toughest science problems you face where you think information technology may be able to help?
2. What are the specific information services (including data/models/ applications/operational products from others) or capabilities you think you need?
3. What results or improvement to the way government serves citizens would exist, that doesn’t now, if you had this service or capability?
Data and Information Technology Questions
1. What capabilities can you provide, or are you aware of, that you think have the greatest potential to help researchers, modelers, and analysts?
2. What new capabilities are needed? What needs to happen to bring these capabilities into existence?
3. What are the most important technical barriers to and gaps in interoperability and how do you solve them?
Cross Domain Communities Questions
1. If I ruled the world, what would I like to have that the community in the other room could offer me?
2. What could I offer the other community that I think would be useful to them?
3. What models and model outputs exist that can benefit both communities that should be coordinated? What are the shared needs (i.e., for certain types of data such as national land coverage, water boundaries etc.)?
4. What are the known interactions between domain communities such as drought related impacts on air quality?
5. What is the shared interest/where are the overlapping needs for services such as geospatial mapping, analysis and visualization of information. What other services could be shared?
6. If we “mashed up” AQ and drought information, what opportunities would that create that we might miss if each domain proceeded separately?
7. What other observations would you like to put on the table?
8. What are the positive and negative implications of sharing services across domains?
11:00 – 12:00 Report outs and discussion 12:00 – Lunch – end of morning sessions
Afternoon Session: Identifying and Building Tools that Cross the AQ and Drought (NIDIS) Domains
1:30 – Welcome, introductions as required, recap, and charge to the group [take the work from the morning sessions, talk about priorities, break into groups and work on what you are interested in, and start building the tool].
1:45 – 2:15 Report outs from AQ group
2:15 – 2:45 Report outs from drought group
2:45 – 3:00 Discussion of priorities, selection of high interest projects
3:00 – 5:00 Smaller Groups go forth and get the projects started
5:00 – reconvene and report out on status; closing, next steps [who carries the project forward, for instance]
5:30/6:00 – close
Contact: Kathy Fontaine, [email protected]
RAW, UNPROCESSED Notes from the session described above. NOTE: We did not reconvene, so there are not closing notes. Please excuse the weird formatting - that'll be fixed later.
Diagram from report
Italics = Drought Normal = Air Quality
Particles-precipitation-radiative effects (link down to direct/indirect effects below) Airborne particles-hitchhikers on the particles Weather forecast model-dust forecast model Dust generation – landcover-forest fires (link down to direct/indirect effects below) Climate change Climate change Human health Air quality-food production-water availability Bacterial content in dust
Direct/indirect effects of particles—feedback (- and +)
Satellite Data Coverage-precision-frequency of images (temporal)
In situ Monitors/Measurements AirNOW Etc.
Other Platforms Ships, airplanes, weather balloons, ground-based lidar
Models Gap filling between monitors Prediction capability “What if?” questions
Decisions Long-term forecasts Alerts & seasonal forecasts Hazard avoidance-immediate health threat
Interactive Datas Proposal: prototype data interoperability for drought & air quality decision making and health impacts
Proposal SOA (service-oriented architecture)
IT & integrating IT services and ontologies Data Model Decision Support
Use Cases Health impacts of pathogents/stressers in dust
Track, analyze, support decisions . . . .
Hemispheric transport of air pollution
Sahel dust, Gobi Fires/biomass burning “Brown Cloud”-Indian Ocean
NIDIS “leaving the data on the portal” is not enough
Stakeholder “consumption perspective”-participatory R&D (continual, “democratic” involvement) (more than “classes” of users)
Past-Present-Future web services to efficiently transform these into one usable . . . (line to linking variables and triggers/alerts below)
“Two clicks is one click too many”
Linking variables (line up to past-present-future bullet above)
Triggers/alerts/monitoring linkages and tie in to (line to past-present-future above)
Need to identify “what happens when a serious situation/drought is trigger
Vulnerability assessment Policy-action
Provide a broader spectrum of data/info products . . . to NIDIS
How can ESIP help NIDIS achieve/select
CVASHI and regional (national) like AHIS
Level of specificity
Trying NOT to reinvent
-metadata -providence -portlets -translators - semantic web
But must make use of underlying body of info
Agree on a common data model
SWEET for Earth and environmental science (semantic web meeting) ontology/vocabulary to be used in a semantic web
NEEDS-where to start?
-With longer term- “Anthropology” approach -Stakeholder impetus – example of how AZ Hydrologic Info Sys got status (data/orgs and how AZ Water Institute got started (policy/linking research management society
IDEAS FOR PROJECTS
1. CLIDDSS (Holly) WCS/DataFED Linking products to show historical trends Giovanni
2. What goes in the box? (Phil) How to architect AirNOW in a box Aeronet
3. Subsetting Calypso/Satellite Data (Skip for now)
4. Links between AQ, drought & public health (John H.)
5. New capabilities for AirNOW (Steve Y.)
6. Emissions Inventories (Stefan) What does it take? What’s there? What’s needed?
7. Potential Projects (Glynis) Sandboxes?
Link AirNOW w/NOAA—weather.gov
Find neutral places of agencies to collaborate
Integrate surf measure. -Satellite data model
Package components/create profiles w/open web services through identifying use case and search for pieces
Visualization tools -Easily understood
Near real time data availability -Funding issue (to provide) -PHaiRS
Global surface AQ data for model verification & forecasting -Lots of networks there, but not connected -Define dist. stds -Similar to met. data? -Data availability issue? -Start w/N.A. database?
Catalogs talk to each other
Ontologies created for each?
International real time data standards -Format mostly -*also metadata -Transport . . . . ALL
Q&A -Latency -Lineage -Feedback to provider
Would enable an AirNOW in a box
Public Health ↔ AQ
Particulate matter -Better resolution @ model level & obs. Advanced Monitoring Initiative -EPA-funded/NE US -NASA funded one for Atlanta
Put applications people on science teams (Decadel Survey rec. also)
Subsetting tool for satellite (EPA AirNOW) data -Emulates OPenDAP
RSIG? (ESIP evaluate?) -Remote Sensing Information Gateway
Interface in a box (ESG) (GEOSS) (data access & analysis) missing piece Svcs in a box Discovery Put WCS on the interface Bet. Data & catalogs
Metadata are inadequate for catalog svcs. -Missing pieces -Lineage (CEOP, WCS opps?) [email protected] parameter level, e.g., gridding assumptions -Metadata file may not be connected to data—embed or refer to -Need self-describing metadata
Reliability -Data dropouts -Stable ftp server for met. forecasts (opportunity)
Emissions (incl. fires) inventories (w/NIDIS?) -Quality of . . . -Architecture of . . . -NEISGEI-data fed -Carbon budget issues? -Blue Sky RAINS -National, global intercomparisons Dust transport
“IDEA-in-a-box”—target China Olympics? -(GEONetCast?) ITAR issue?