HR: 1340h
AN: H13A-0978    [Abstracts]
TI: Lessons Learned from the Deployment of a Hydrologic Science Observations Data Model
AU: * Beran, B
EM: bora.beran@gmail.com
AF: Microsoft Research, 455 Market St., Suite 1690, San Francisco, CA 94105, United States
AU: Valentine, D
EM: valentin@sdsc.edu
AF: San Diego Supercomputer Center, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093, United States
AU: Zaslavsky, I
EM: zaslavsk@sdsc.edu
AF: San Diego Supercomputer Center, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093, United States
AU: van Ingen, C
EM: vaningen@windows.microsoft.com
AF: Microsoft Research, 455 Market St., Suite 1690, San Francisco, CA 94105, United States
AB: The CUAHSI Hydrologic Information System project is developing information technology infrastructure to support hydrologic science. The CUAHSI Observations Data Model (ODM) is a data model to store hydrologic observations data in a system designed to optimize data retrieval for integrated analysis of information collected by multiple investigators. The ODM v1, provides a distinct view into what information the community has determined is important to store, and what data views the community. As we began to work with ODM v1, we discovered the problem with the approach of tightly linking the community views of data to the database model. Design decisions for ODM v1 hindered the ability to utilize the datamodel as an aggregated information catalog need for the cyberinfrastructure. Different development groups had different approaches to populating the datamodel, and handling the complexity. The approaches varied from populating the ODM with a bare minimum of constraints to creating a fully constrained datamodel. This made the integration of different tools, difficult. In the end, we decided to utilize the fully populate model which ensure maximum compatibility with the data sources. Groups also discovered that while the data model central concept was optimized for data retrieval of individual observation. In practice, the concept of data series is better to manage data, yet there is no link between data series and data value in ODM v1. We are beginning to develop ODM v2 as a series of profiles. By utilizing profiles, we intend to make the core information model smaller, more manageable, and simpler to understand and populate. We intend to keep the community semantics, improve the linkages between data series and data values, and enhance data discovery for the CUAHSI cyberinfrastructure.
DE: 0525 Data management
DE: 1805 Computational hydrology
DE: 1819 Geographic Information Systems (GIS)
DE: 6339 System design
SC: Hydrology [H]
MN: 2007 Fall Meeting