A Virtual Observatory for the Ionosphere-Mesosphere-Thermosphere Community

1. A Virtual Observatory for the Ionosphere-Mesosphere-Thermosphere Community D. Morrison, M. Weiss, R. Daley, L. Immer, S. Nylund, J.-H. Yee, and E. Talaat (Johns Hopkins University, Applied Physics Laboratory 11100 Johns Hopkins Rd. Laurel MD, 20723: 240-228-4172) J. Russell (Hampton University) R. Heelis (Univ. Texas at Dallas) J. Kozyra (Univ. of Michigan) D. Bilitza, R. McGuire, and R. Candey (NASA Goddard Spaceflight Center) P. Fox (HAO/NCAR) VxO Workshop

2. ITM is Diverse in Observables An ITM VO must facilitate correlative study between multiple parameters from multiple sources and cross-discipline studies. VxO Workshop

3. VITMO Constituency • VITMO is focused on the “interdisciplinary” ITM user rather than the “instrument team” user. • Focus is on facilitating multi-instrument and multi-mission investigations. • Provide tools and capabilities to tie together disparate measurements of the same or related phenomena. VxO Workshop

4. People, Institutions, Data Sets • D. Morrison – Principal Investigator • M. Weiss, R. Daley, L. Immer, - SRAS Principals • Science Advisory Panel • J.-H. Yee and E. Talaat - APL • J. Russell - Hampton University • R. Heelis - Univ. Texas at Dallas • J. Kozyra - Univ. of Michigan • D. Bilitza, R. McGuire, and R. Candey - NASA Goddard Spaceflight Center / Raytheon • P. Fox - HAO/NCAR • Data / System Providers • D. Morrison – APL – GUVI/TIMED, SSUSI/DMSP, AURORA/NPOESS • S. Nylund – TIMED MDC Integration – APL • J. Russell - Hampton University – SABER, AIM, UARS • R. Heelis - Univ. Texas at Dallas – C/NOFS, DMSP SSIES • D. Bilitza, R. McGuire, and R. Candey - NASA Goddard Spaceflight Center / Raytheon – SPDF, CDAWeb, ModelWeb • P. Fox - HAO/NCAR – CEDAR, VSTO • Scott Bailey – Univ. Alaska – SNOE • Jan Merka - NASA Goddard Spaceflight Center - VMO VxO Workshop

5. The Ability to Find the Appropriate Data Will Be a Key Requirement of a VITMO Increasingly satellites in the ITM community will rely on remote sensing instruments – TIMED, DMSP, NPOESS VxO Workshop

6. How the VITMO Will Be Different • For a VxO to be truly useful it must go beyond traditional mission data centers. • It should be able to interrogate multiple data centers, finding products that overlap in time and/or location. • It must allow a higher level of search to be performed than previously available. An example of such a complex search would be something like the following: “what near Earth orbiters are observing the auroral region when Bz is negative and the solar wind speed is greater than 400 km / sec.?” Today, this would require a manual process where the researcher would first find plots of the Earth’s magnetic field component (Bz) and plots of solar wind speed from NASA’s CDAWeb (http://cdaweb.nasa.gsfc.gov). After they manually looked for overlap conditions they would utilize these time periods to search known Earth orbiters to find out if they were observing during those time periods. They would then have to manually review summary products to determine if those orbiters were viewing the auroral region during those intervals. VITMO will handle these types of queries automatically. • VITMO will also be able to organize tools, whether plotting, subsetting, or analysis tools by the type of data they are to be applied to as well as the types of operations that are to be performed. If the user requested time series data then tools appropriate for operating on the time series should be presented to them. If they chose images then tools appropriate for images should be made available. If the data were in CDF format, then FITS format data tools need not be presented. Additionally, VITMO will understand that model output can be treated as high level data products and that models should be available, just as tools are, to the end user. VxO Workshop

7. VITMO is Java based. The Knowledge Base is currently an Oracle database. VxO Workshop

8. VITMO Architecture Overview • high-level conceptual model of the scientific disciplines and domain data in the region from the Sun to the Earth • Regions – Solar, Near-Earth – F-region, E-region • Features – exist for long periods of time • Events – Occur at particular points in time • description of resource metadata in the Knowledge Base supporting any accessible scientific resource • descriptive metadata includes program, satellite and instrument information, as well as data details like observed phenomena, units of measure, time and spatial coverage, processing level • structural metadata defines the details for data access and retrieval if selected by the use • expand the notion of a data resource in our metadata definition to also include databases and computer programs (such as format conversion tools or data models) that generate requested data on demand. Thus, file access, database access, and program invocation details are all included in the resource metadata model. VxO Workshop

9. Coupling in VITMO Architecture • a loose coupling between the conceptual model and the resource metadata. • A query in the conceptual model identifies the desired parameters, which are then mapped to the related data resources. Data discovery is simplified by performing requests at the conceptual level, permitting users without knowledge of specific mission, or instrument to readily find data of interest. • conceptual model can be readily expanded if new classes of scientific concepts or new relationships between scientific concept classes would improve the model's representation of the Sun to Earth domain. • a data provider can add resources (of any format, access method, or processing level) simply by providing information that describes the data and its access details. • model eliminates the need for rigid metadata standards or unique, resource-specific integration software VxO Workshop

10. Scientific Conceptual Model in the VITMO Knowledge Base VxO Workshop

11. Descriptive metadata portion of the VITMO Knowledge Base VxO Workshop

12. Structural metadata portion of the VITMO Knowledge Base VxO Workshop

13. VITMO Based on Services VxO Workshop

14. Enhanced Search Services • Today’s researcher needs to find datasets from different missions that overlap in time and space. • Many of today’s satellite missions include remote sensing instruments. Finding cross comparisons between instruments is more difficult because we must be concerned not just with spacecraft location, but also with multiple look-points for the instruments. • The prime NASA mission for study of the MLT region is the TIMED satellite. All of the instruments on this satellite are remote sensing instruments. Each remote sensing instrument has a different look geometry, some looking as far as 2000 km away from the satellite. VxO Workshop

15. Enhanced Search Services Continued • AIM will have a fly’s eye imager. • The fields-of-view (regard) of these instruments define the extent of the data products. • This location information is not contained in the existing metadata in today’s catalogs. VxO Workshop

16. Virtual Metadata Generators • VITMO will employ special coincidence calculators that know the viewing geometries of the various instruments. • The TIMED coincidence calculator is an example of a tool which when converted into a web service can be used to augment existing data querying. • The TIMED coincidence calculator has recently been imbedded into the TIMED query system. • These calculators will operate as web services inside the VITMO query system (or by outside query systems). • These calculators generate metadata for the coincidence of interest which is used by the catalog system to answer your query. • The metadata generated by this service is discarded after it is used – hence “virtual metadata”. VxO Workshop

17. Enhanced Search Results • Addition of virtual metadata capabilities allows pre-selection of data for analysis. • VITMO search will deliver results from searches such as “What GUVI data that contains 23-01 MLT during active geomagnetic conditions?” VxO Workshop

18. Integrating with VITMO • VITMO can integrate with data providers directly – understanding provider data and file structure. • VITMO can integrate with high level providers (other VxOs) through services (i.e. the way we will integrate CDAWeb). • VITMO can integrate with high level query services and data providers (and older style providers) through Front Door Integration – typically how the VITMO could interface to NGDC and SPIDR. VxO Workshop

19. The following is an example of mapping VITMO parameters to another system, i.e. SPDML. Front Door Integration • We have prototyped and developed a new approach that can allow many of the capabilities of an existing data center to be easily incorporated into the VITMO by interfacing to their system by a specialized web service. • Consider, for example, a data center that provides data query and retrieval services. A significant amount of time and effort will certainly have been spent developing the user interface to such a system. • We have developed services that encapsulate that entire user interface, allowing user input, enabling a combined system to take full advantage of the rich capabilities that already exist. • We demonstrated the feasibility of this approach by incorporating the Space Physics Data Markup Language search engine (SPDML) into the SRAS system (VITMO predecessor system). • Advantages • Allows reuse of user interfaces • Concept mapping is high level (user level) rather than at the database or metadata level – much simpler. • Allows integration of extensive search services. VxO Workshop

20. Prototype Display The left hand side of the web page allows the user to search or browse for resources (be they data, models, tools, etc) by the particular region of space, time, satellite coincidences or other criteria. After resources are searched for the results are displayed and possible display plots may be shown. If the user selects the “Related Tools and Models” button on the top of the page a display of appropriate tools and models for the data or resources chosen will be presented. This list of tools and models is generated dynamically based on the metadata in the VITMO catalog describing the data, tools, and models available through the VITMO. VxO Workshop

21. Schedule • Year 1 • Perform a preliminary task analysis for ITM data query. This will help define the number of the web services that need to be created and establish the parameters that need to be passed. We will define the lists of rules that will be used to determine what web services are called and the order that they are called to satisfy the data query. • Work on defining the WSDL that will be used to pass messages between the different web services. • Define integration approach for web service capabilities provided by CDAWeb and SSCWeb. Define any changes to existing web services they provide. Define additional web services desired. • Define model and tool metadata. Evaluate MODELWeb models as services. • Develop general-purpose mechanisms to incorporate web service virtual metadata generators into the data discovery request process. • Survey all data products, tools, and models available for inclusion in first build of VITMO. • Present results at an AGU meeting and/or an NSF CEDAR meeting. • Year 2 • Continue implementing the rule-based middleware used to determine the number and order of query operations. • Develop web interface based on feedback from .previous workshop. • Implement metadata describing tools and models into VITMO. • Implement the TIMED solar geophysical database as a web service. • Provide a workshop to solicit community input and demonstrate proposed approaches and solutions to integration problems and search capabilities. • Implement MODELWeb models as web services. • Present results at an AGU meeting and/or an NSF CEDAR meeting. VxO Workshop

22. Schedule – Year 3 • The third year will focus on tying all of the pieces together into a complete demonstration system. • We will have TIMED, UARS, SNOE, C/NOFS, AIM, Polar, IMAGE and ACE data sets (available at SPDF [Bilitza, 2003]), DMSP from F15, F16, and F17 on. In addition, we will have a selection of ground sites through VSTO, as well as the SuperDARN radar network, added in. • We will tie our web services architecture into our existing VITMO system which will already have an interface for making the type of queries we require. • Integrate with other VxOs (VHO, VSO, and VMO) to allow searches between systems and sharing of tools, models, etc. • Provide a final workshop to solicit community input and demonstrate proposed approaches and solutions to integration problems and search capabilities. Propose remaining problems in the development of a complete operational system (security, performance, etc). • Present results from workshop at an AGU meeting and/or NSF CEDAR meeting. • Develop a guidebook describing how future missions can become part of the VITMO. • Publish software developed and web services protocols used. This will allow outside organizations such as NSSDC, NGDC, and future virtual observatories to tie into our system to enhance their existing capabilities. VxO Workshop

23. Community Input • Routine input from Science Advisory Panel • Annual workshops demonstrating functionality, refining requirements, etc. • AGU – demonstrations • First year input (including Science Advisory Panel) may be postponed to second year to provide more funding for software development (strong review panel recommendation). VxO Workshop