5 A Framework for Automated Forecast Verification and Dissemination of Performance Information

Monday, 7 January 2013
Exhibit Hall 3 (Austin Convention Center)
Paul Hamer, CIRA/Colorado State Univ., Boulder, CO; and M. A. Petty and J. L. Mahoney

Handout (1.0 MB)

The Forecast Impact and Quality Assessment Section (FIQAS) of the Global Systems Division (GSD) is responsible for the development of verification techniques and technologies that allow for the generation and dissemination of statistical information on forecast performance in near real time. In order to achieve this goal, the components of forecast verification have been captured in a framework to support the configurable deployment of solutions according to the roles required for verification, be it feedback to the forecast process, incorporation into decision support tools, or monitoring of forecast performance.

The forecast verification problem has been decomposed into three distinct layers in an architecture that handles:

1. Production: the acquisition of data, event-driven processing and pluggable verification components;

2. Integration: the capture of verification result sets and metadata;

3. Presentation: the ability to query and disseminate statistical information for analysis;

This framework has been applied in varying degrees to technologies within FIQAS, and has been developed using open source software and tools used in the development of NextGen to more easily support the integration of verification into real-time forecasting and decision-making processes. Further details of the framework components along with FIQAS technologies using this framework will be discussed.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner