The Weather Information Statistical Post Processing System (WISPS): A community-based statistical post-processing system

The Meteorological Development Laboratory (MDL) of the National Weather Service (NWS) has begun to design the Weather Information Statistical Post-Processing System (WISPS), a statistical post-processing (StatPP) software system for numerical weather prediction data. The new software framework will be built to facilitate modern collaboration, to leverage high performance computing, and to embrace widely-accepted standards for encoding data and metadata. By using numerical libraries for rapid computation, modular software design for maximum collaboration, and data formats based on NetCDF, we hope WISPS will become the new standard for StatPP software in the global weather enterprise.

  The vision of the WISPS project is to create an open-source, community-available software system that also can facilitate a more rapid Research-to-Operations process for a variety of StatPP techniques. WISPS will use tools such as NOAA’s Virtual Laboratory (VLab), a software repository, community forums, and modern project management, software integration, and bug tracking software. WISPS standards for data and metadata encoding will utilize ISO and WMO standards, and the Climate and Forecast conventions, and online code registries. Finally, WISPS will be highly documented, both internally in the software and externally with a user guide and FAQ, and technical and scientific documentation.  Initial software development has focused on the operational StatPP needs of the NWS, and it has used Python.

  The extensibility, flexibility, and shared nature of WISPS positions it to meet the evolving needs of both NOAA and the weather enterprise.  Software structure, data, and metadata concepts for StatPP applications are presented.