A NoSQL storage design for meteorological grid data products is introduced. An analysis is made of the meteorological structure of grid data products and their application scenarios, and the column storage database in NoSQL is designed to increase the efficiency of storage and access through optimizing the way of table store. According to the type, time and forecast effectiveness of data products, we work out the primary key column, attribute column, and data column, then, split the big grids of data column, use the grid sequence number to record data column names, and design data compression strategies to improve the overall efficiency of data access. Taking numerical prediction products as an example, we make use of Aliyun OTS to simulate tests. The results show that the column storage database based on NoSQL technology can meet with the requirements of the underlying data support of product service platform, which has a certain reference significance for the extension and application of key technologies, such as distributed data environment, cloud computing, and so on.