SURA Coastal Ocean Observing and Prediction Program

The Southeastern Coastal Ocean Observing and Prediction (SCOOP) Program is an ongoing collaboration between the modeling research community and operational agencies, such as the National Oceanic and Atmospheric Administration (NOAA). SCOOP, sponsored by the Office of Naval Research (ONR) and NOAA's Coastal Services Center, is a multi-institution collaboration which aims to take today's cutting-edge activities from the research community, and develop these so they can form the basis for tomorrow's operational systems. The architecture of the SCOOP Archive, showing the various components, and how they interact.

Part of this project's work involves the regular execution of atmoshperic (e.g. WindGen) and hydrodynamic (e.g. ADCIRC, WaveWatch3) coastal modeling codes, for various geographical regions.

Additional runs of some codes are performed to predict the path and effects of ongoing tropical storms and hurricanes; the results from these runs are passed to groups involved in evacuation planning. Specifically, resources at CCT are used by the Bedford Institute of Oceanography to run WaveWatch3 on ensemble predicted hurricane tracks sent out by the University of Florida in response to Hurricane tracks produced by the National Hurricane Center. Outputs from these runs are sent out to other project partners for visualization purposes. We are nearing the point at which this data pipeline operates without any human intervention.

To support this, the Center for Computation & Technology at LSU operates an experimental archive of data for use by the SCOOP Partners This archive stores atmospheric model outputs (wind data), results generated by the hydrodynamic models, which use the atmospheric model outputs for input (wave/surge data), and observational data to be used for verification of model results (sensor data). Unlike the production archive at TAMU, the focus of the LSU Archive is to use emerging technologies, such as Grid Computing, to provide easy and secure access to this data. The archive forms a backbone for the research efforts of the project, and as such, has to be both highly available, and reliable. Although supporting SCOOP was our prime objective, we wanted to be able to re-use most of the archive's functionality, and code, for other efforts with data storage requirements, such as CCT's numerical relativity work. The archive is based around a 5TB local storage area, plus a 7TB allocation of off-site storage at SDSC's DataCentral. The goal of the archive is to provide easy access to this data, no matter where it is stored.

As an initial demonstration of the archive, in 2005, a portal was constructed using the GridSphere portal framework. This portal remains in use, and provides a number of facilities to users. These include the ability to search the archive for files, which can then be downloaded using either http or GridFTP, to the user's desktop, or directly to a remote location (GridFTP only). The portal also provides a simple interface that allows popular coastal modeling codes, e.g. ADCIRC, to be run and tracked on the SCOOPGrid testbed resources; a monitoring page informs users of the availability of these resources. This is currently being extended to support a wider range of codes.

We have also designed and implemented command-line tools that provide access to the archive files. GetData, which was built using the Grid Application Toolkit (GAT), supports file downloads using http and GridFTP, and provides an interactive query mode where the user can enter search parameters. MiniGetData (mgd) is a cut-down version of GetData, built on top of wget, allowing http downloads only. Both tools are in the process of being deployed to the SCOOP Partners.

Participants

Senior Staff

• Dr. Gabrielle Allen (CCT/CS)
• Dr. Jon MacLaren (LSU Project Manager)
• Brett Estrade
• Archit Kulesthra
• Shree Balasubramanian
  

Graduate Students

• Chirag Dekate
• Dayong Huang
• Theresa Xu

Contact
For more information about the SCOOP project at CCT, please contact Dr. Jon MacLaren.

Acknowledgements
This program is a multi institution collaboration sponsored by the Office of Naval Research and NOAA's Coastal Services Center .