Rivers and streams are among the most endangered ecosystems on Earth, with rapid population declines due to habitat degradation and water pollution. The recent proliferation of the use of in-situ sensors in streams and rivers supports space-time water quality modelling and monitoring in near real-time. In this presentation, we introduce a new family of spatio-temporal models for river networks, in which spatial dependence is established based on stream distance and flow connectivity, and temporal autocorrelation is incorporated using vector autoregression approaches. A framework for anomaly detection is also developed using these spatio-temporal models. The approaches are illustrated using two case studies:  water temperature data in the northwestern United States and new sensor data in streams in Queensland Australia.