Cyberinfrastructure to support Real-time, End-to-End, High Resolution, Localized Forecasting

by

Doug Lindholm, Tom Baltzer, Mohan Ramamurthy, and Ben Dominico

 

Abstract:

 

From natural disasters such as flooding and forest fires to man-made disasters such as toxic gas releases, the impact of weather-influenced severe events on society can be profound. Understanding, predicting, and mitigating such local, mesoscale events calls for a cyberinfrastructure to integrate multidisciplinary data, tools, and services as well as the capability to generate and use high resolution data (such as wind and precipitation) from localized models. The need for such end to end systems -- including data collection, distribution, integration, assimilation, regionalized mesoscale modeling, analysis, and visualization -- has been realized to some extent in many academic and quasi-operational environments, especially for atmospheric sciences data.  However, many challenges still remain in the integration and synthesis of data from multiple sources and the development of interoperable data systems and services across those disciplines.

 

Over the years, the Unidata Program Center has developed several tools that have either directly or indirectly facilitated these local modeling activities.  For example, the community is using Unidata technologies such as the Internet Data Distribution (IDD) system, Local Data Manger (LDM), decoders, netCDF libraries, Thematic Realtime Environmental Distributed Data Services (THREDDS), and the Integrated Data Viewer (IDV) in their real-time prediction efforts.  In essence, these technologies for data reception and processing, local and remote access, cataloging, and analysis and visualization coupled with technologies from others in the community are becoming the foundation of a cyberinfrastructure to support an end-to-end regional forecasting system.  To build on these capabilities, the Unidata Program Center is pleased to be a significant contributor to the Linked Environments for Atmospheric Discovery (LEAD) project, a NSF-funded multi-institutional large Information Technology Research effort. The goal of LEAD is to create an integrated and scalable framework for identifying, accessing, preparing, assimilating, predicting, managing, analyzing, mining, and visualizing a broad array of meteorological data and model output, independent of format and physical location. To that end, LEAD will create a series of interconnected, heterogeneous Grid environments to provide a complete framework for mesoscale research, including a set of integrated Grid and Web Services.

 

This talk will focus on the transition from today’s end-to-end systems into the types of systems that the LEAD project envisions and the multidisciplinary research problems they will enable.