It is important to establish a long-term monitoring network and a platform for collecting and managing observation data for agriculture in China. Combining with the construction practice of the National Agricultural Science Data Center in recent years, this paper systematically combs the observation-data aggregation process of the long-term monitoring network of agriculture in China; gives the standard norms involved in data aggregation management; and designs and implements a platform for the management of the aggregation of agricultural science observation data, which comprises an agricultural observation data acquisition and processing system, agricultural observation data mining analysis and early-warning system,agricultural observation data exchange system, agricultural observation data long-term preservation system, and agricultural observation data sharing service system. These systems have three functions: long-term preservation of agricultural scientific observation data, data sharing, data analysis, and mining application. The preliminary application of the platform shows that the design of the structure and function of the platform is reasonable, which provides solid data support for agricultural scientific research and agricultural macroscale management decision-making. The platform can be used for reference and guidance in similar research and application.