Constructing a Web Service System for Large-Scale Meteorological Grid Data

In this paper we report a Web service system constructed on top of the GPV/JMA Archive. The archive is to offer the daily operational weather forecasting data provided by the Japan meteorological agency (JMA). Currently, it is serving several kinds of GPV data, such as global spectral model data, regional spectral model data, and meso- scale non-hydrostatic model data. In order to make those data more usable from the Net, we have constructed a Web service system, consisting of several Web services, such as metadata generation, searching, and rendering, and a workflow engine that allows users to execute BPEL based scientific workflows. In particular, we have designed a dedicated metadata format for GPV data, and implemented metadata extractor to enable automatic metadata extraction for incoming GPV data. In this paper, the entire system and the details of the constructed services are described.