Linked Environments for Atmospheric Discovery (LEAD) is a multi-institutional Large Information Technology Research project funded by the National Science Foundation (NSF). The goal of LEAD is to create a framework based on Grid and Web Services to support mesoscale meteorology research and education via provision of capabilities such as running of steered forecast models, data mining for meteorological phenomena, and dynamic workflows that automatically reconfigure themselves in response to changing weather. LEAD presents unique challenges in managing and storing large data volumes from real-time observational systems as well as data that are dynamically created during the execution of adaptive workflows.

At IIPS 2006 in Atlanta, an architecture for the THREDDS Data Repository (TDR) was presented. This architecture uses a facade pattern to provide the programmatic user of the TDR access to storage repository capabilities through two simple functions of putData() and getData(). Behind the facade, important functions of locating storage for the data, moving the data, assigning unique identifiers, generating metadata and performing name resolution are performed using implementations suitable for a particular application.

The LEAD Data Repository (LDR) is an implementation of the TDR currently under development that provides a collection of services to support the large-scale data storage and management needs of the LEAD project. Via a web service interface it must be able to autonomously handle storage and retrieval requests generated by LEAD workflows in addition to directly satisfying user requests. This includes generation of, access to, and management of metadata that supports discovery and use of the data. In particular, the LDR must integrate with the THREDDS Data Server (TDS), which in turn provides middleware for metadata access and management, OPeNDAP access, and integration with the Unidata Integrated Data Viewer (IDV), among other benefits.

This paper will summarize TDR and LDR development and capabilities and integration into the LEAD project.