Design And Implementation Of The Data Retrieval And Visualization System For Tsunami Warning System

Tsunamis are giant waves caused by earthquakes, volcanic eruptions under the sea, submarine landslide or other disturbances. Tsunamis are often destructive and trigger other hazards, therefore, it is significant to warn tsunamis timely and accurately. CTSU (Tsunami Numerical Model) is applied in tsunami early warning now, and produces about 400million logs and uses 20TB storage on supercomputers. Thus it is critical to build a new tsunami warning system to realize quickly retrieving and processing of multi-source data, efficiently storing, searching and visualizing of mass tsunami scene data.The main achievements of this thesis are as follows:(1) Current studies and developments about methods of mass data retrieval and visualization and worldwide tsunami warning systems are reviewed; Techniques of mass data storage, multi-source data collection and process and visualized analysis based on 2D/3D GIS (Global Information System) are studied.(2) The requirements of mass data searching and visualizing are analyzed, including the requirement to design the whole system and requirement of the function and performance; The principles of system design is introduced, and system scheme design and the decompose of basic function modules are carried on accordingly.(3) The key module to searching and visualizing mass data are designed and implemented, such as data storage and retrieval modules; data collection and process modules; the design of UI, algorithms and implementation of tsunami scene timely analysis and visualization modules.(4) System deployment of mass data retrieval and visualization is designed:Pilot run of the tsunami warning system in several business units is introduced; Tests of system function and performance is given in the thesis:The performances of several parts of the system are shown and analyzed.The Tsunami Warning System Platform has been built, tested and used in systematic operation in National business units. The system performance so far is good and can preliminarily satisfy the need of tsunami warning relative businesses. Earthquakes and Tsunamis data collection and process, mass tsunami scene data searching and GIS-based visualized analysis modules can satisfy the functional and performance requirements of architecture design. The future study will focus on continuously optimize data storage and retrieval methods due to the increase of tsunami scene data, in order to further improve the reliability and utility of the tsunami warning system.