Abstract: Multi-source mobility data provides an opportunity to understand urban phenomena and enjoy the urban life. Existing efforts have been made for studying multi-source mobility data via statistical physics models and data mining technologies. However, little attention is paid on visual analytics methods. In this paper, a visual query model is proposed to study multi-source human mobility data in a city to handle challenges from data management and visualization. First, principles of data abstraction are defined to produce a general mobility data model. Second, several novel data management methods are employed to enhance fetching data in real-time. Then, a query interface and a series of visualization techniques are introduced to improve user's analysis. Finally, a demo system is implemented and some typical cases demonstrate the effectiveness and efficiency of our model.