P4.2
The Real-Time Verification System (RTVS-NG): Integrating data to support user-based verification
Nick D. Matheson, OAR, Boulder, CO; and M. A. Petty, D. S. Schaffer, S. Madine, and J. Mahoney
The Real-Time Verification System (RTVS-NG): Integrating data to support user-based verification
The En-route Verification Proof of Concept (ERVPOC) is an initial implementation of the GSD Real Time Verification System Re-engineering (RTVSNG) effort and the first realization of the Verification Framework concept. The goal of the RTVSNG effort is to provide an updated verification system that meets user requirements in an adaptable and agile manner. Exploring the breadth and depth of requirements for the RTVSNG project emphasized the need for integration, comparison, analysis and dissemination of verification data from numerous sources across organization boundaries. The primary challenge is how to provide this type of integrated data analysis environment without imposing unnecessary or unrealistic constraints on the sources and organizations involved. The belief that a Verification Framework is needed grew out of this challenge. A Verification Framework will unify these sources and organizations, not at the level of software, tools and file formats, but rather at the level of data sharing mechanisms, agreed upon within the verification community, that will ultimately provide a more complete assessment of the quality and value of the products being verified. Once a Verification Framework is in place, Software Architectures and Software Realizations can be developed and evolve without needing to reinvent the Verification Framework. The ERVPOC is one example of a Software Realization within the Verification Framework.
One important role of the Verification Framework is to facilitate sharing of data in the verification community. This will enable the combining of in-house verification results with products from external verification systems and tools, as well as allow extraction of products by individual users to perform their own analyses. As a result, the Verification Framework will expose data of various types, such as verified forecasts, observational data, and user-based data, from a diverse range of sources. These data can be represented at many levels of detail and at various stages of analysis. Relevant attributes of these datasets will be used to describe dataset Vectors, which in turn will be the mechanism for import and export.
Another key role of the Verification Framework is to facilitate meaningful associations amongst datasets. These associations are dictated by user domains, examples of which are the comparison of competing forecast products, or the analysis of the impact of a forecast product's performance intended for decision support, as in the case of traffic flow management. Dataset Vectors needed for a comparison may not have attributes directly in common with the user domain, and will warrant an intermediate mapping to serve as a bridge from the dataset's native Vectors to the common space of the user domain.
Providing verification data analysis by integrating convective forecasts from different sources and of different types (object based, dichotomous and probabilistic), combined with air traffic cost measures presented in the context of air traffic decision making, is the core business case for the ERVPOC. Four focal areas have been developed from this business case:
The ERVPOC business cases are achieved via a Software Architecture and a subsequent Software Realization that contain a core Verification Statistics Server (VSS) that is leveraged by focused User Interface implementations to meet the objectives of specific user groups. The design decouples delivery interfaces, mechanisms and formats for specific user groups while maintaining a shared core for Vector management, mapping, joining and analysis.
The ERVPOC is the first tangible demonstration of not only the core concepts of the RTVSNG effort, but also the relationship between the Verification Framework, a Software Architecture and a Software Realization. Furthermore, this was accomplished while employing new verification, air traffic measurement and user centered analysis techniques. The project has provided its share of learning experiences and challenges. Areas for future investigation and work have been proposed from these successes, failures and challenges.
Poster Session 4, Radar and Icing Posters
Thursday, 24 January 2008, 9:45 AM-11:00 AM, Exhibit Hall B
Previous paper Next paper