The invention discloses a distributed storage method for massive Argo data. The method can be used for achieving high-efficient storage of multi-source heterogeneous, dynamic and multidimensional and massive Argo data. The method comprises the following steps: 1) a reasonable table structure is designed, 2) different data organizations and management modes are used according to different characteristics and application demands of the Argo data, 3) based on HDFS, distributed storage of the massive Argo data is realized and automatic load balancing is further realized, in data transmission, a system uses the technology of combining HDFS multiple nodes and virtual IP to solve the problem that multiple nodes cannot communicate with the outside simultaneously; 4) the massive Argo data are automatically stored in distributed cloud storage. The distributed storage method for the massive Argo data disclosed by the invention has important practical application values in Argo data management, and is further wider in application prospects.