Abstract: Ground wireless sensor network is an important method for large-scale observation and information collection and how to scientifically and efficiently manage the big volume of data is one of key issues in this field. In this study, Oracle 10g tool was used to build the observation database and the Microsoft Visual Studio.net, ESRI ArcGIS Engine 9.3 and ERDAS 9.2 software were integrated together to develop the database management system, which allows to seamlessly collect with the wireless sensor network and to automatically receive and manage the data. This system is composed of 4 major modules, i.e., data storage module, data edition module, data browsing and statistical analysis module, data application module. The data storage module is to manage the primary data obtained by wireless sensing network as well as database, including automatically receiving control, data adjustment and error correction, exceptional reports, the database edition module is responsible for creating and maintaining metadata, data retrieval, demonstration and analysis. Basic data statistics, charting analysis and out-put is implemented in the data browsing and statistical analysis module, while the main function of data application module is to realize the spatialization of observed data, spatial-temporal coupled simulation and operation alapplication for agricultural monitoring. This system was tested for one year in Hebi city, Henan Province, which showed that the system is quite stable, secure, functional and convenient in man-machine connection and it can realize the functions of observation data automatic storage, data correlation and processing, and multiple application.