Weather support to consequence management during the 2004 Summer Olympic Games

John Hannan¹ and Patrick Hayes¹

  ABSTRACT             The 2004 Summer Olympic Games in Athens, Greece represented a potential high-value target for disruption by terrorist individuals and organizations.   As part of security operations for the Olympics, an international team of consequence management (CM) analysts employed models that predicted the atmospheric transport and dispersion (AT&D) of chemical, biological, radiological, nuclear, and high-explosive (CBRNE) agents.  One such model, the Defense Threat Reduction Agency (DTRA) Hazard Prediction and Assessment Capability (HPAC) was used by both forward-placed military and civilian personnel in Europe, and by reachback support teams in the U.S.  As is the case with all AT&D models, HPAC requires accurate meteorological data to generate realistic results.

            An Olympic Weather Operations cell was established at the US DTRA in Alexandria to provide 24/7 service to agencies that used the weather forecast data to run CM models.  The weather cell performed two main functions: 1) determine the best performing model forecasts for the time period and host that forecast on the Olympic weather data web page, and 2) provide real-time help-desk support to users who had questions about acquisition and use of weather for CM operations.  Typical guidance provided by the weather ops cell included current status of the weather servers, data cataloguing, which forecast model was working better at the time, and proper HPAC weather configuration for specific scenarios.

Weather forecast data came from three classes of weather models; coarse-resolution, long-range global-scale models such as the NOGAPS and the GFS; high-resolution, medium range mesoscale models such as the MM5, RAMS, and OMEGA models; and the high-resolution, short-range Real Time Four Dimensional Data Assimilation (RTFDDA) system from National Center for Atmospheric Research (NCAR).  Users accessed weather forecast data from classified and unclassified servers hosted by the Technology Directorate of US DTRA in Alexandria VA, USA.  The data were also made available through a suite of operational Meteorological Data Servers (MDS) hosted by DTRA.  The MDS servers were employed to avoid single point of failure issues and were used strictly in a backup fashion.

            The performance of forecast models (accuracy) and associated ensemble uncertainty was measured throughout the Olympic Games.  Performance statistics for the MM5, RAMS, and OMEGA models are shown.

            Consequence management operators provided feedback on how weather affected their operations and how well the Weather Operations Cell supported CM efforts.  Overall evaluation and specific comments from users are provided.  This effort provides good direction for how to establish and operate CM weather support operations in the future.  Several suggested improvements can be made; those recommendations are given.  

¹Northrop Grumman Information Technology, Defense Enterprise Solutions

6940 South Kings Highway, Alexandria VA 22310