Environmental protection directly affects the development of resources, environment and economy, and the protection of the water resources is one part of the primary coverage. In China, the water pollution is very serious, about 80% of the rivers are suffering sever pollution. On the basis of the real-time dynamic monitoring of a large part of natural water resources, this project researches the network-based theoretic approach and technology of multisensory management and data fusion, and constructs the architecture of a data fusion system for monitoring the state of the water. First, considering the performance of sensors, the system computing ability, the communication bandwidth, the expected accuracy, the disposable funds and so on, we present the multisensory management mode, the coordinated control policy, and the relevant intelligent algorithm. Then we complete the simulation test for the original process system for data fusion of the water monitoring system, and the simulation experiments on data-fusion algorithm. Finally, associating with the actual project, we present the approaches of data acquisition and screening, and the remote reconfiguration technologies for the sensors’ operate modes and running parameters.