Recently, more and more scientific simulation fields have generated large-scale three-dimensional (3D) scattered data. To observe the internal information, an interactive visualization technique for extracting the internal information of such data is always needed. However, due to the highly complex distribution and structure of the data, interactive visualization of such data is very challenging. Moreover, the simulation result is also time varying, which makes the visualization more difficult. In this work, we focus on large-scale 3D scattered data, which are generated from a tsunami (tidal wave) simulation. We visualize the large-scale 3D scattered data by developing an interactive particle-based rendering (IPBR) method. A main feature of the proposed IPBR method is that it implements a semi-transparent effect using the ensemble average of particles so that the volume rendering effect, used to observe the internal information of the 3D scattered data, can be realized. Moreover, the real-time adjustment of the transfer function is also implemented by changing the particle radius and particle color. We calculate the density of the 3D point data and adjust the point radius by using a resizing function. Furthermore, we also develop a time-varying rendering mechanism to render the temporal behaviors of the time-varying data. After application of the proposed technique to tsunami simulation data, the experimental results show that a detailed analysis of the spatial and temporal features of a tsunami can be achieved.