At present, database systems such as Oracle and SQL Server are a preliminary application in automatic weather station data storage. The database performance is easy to maintain with a small number of AWS’s, but at the regional level or provincial level meteorological service, because of the large amount of AWS’s and the growing amount of data, the design and implementation of data storage is need to be paid great attention. In 2009, based on the Oracle 10 G database system, the Yangtze River Delta region AWS Database System was developed and established. The system carries out the processing and storing of AWS’s data in real time. The design of the massive database implementation, including event driven storage processing, large table partition, index, and SQL statement query optimization principles in the massive database, is discussed.