中文摘要：

洪水溃坝等复杂灾害现象的时空过程模拟分析，是当前国际地理信息科学热点研究方向之一。随着网络的普及化，公众对GIS时空分析模型的网络实时集成和可视化分析提出了迫切的需求。本文以溃坝洪水为例开展了时空过程网络可视化模拟与分析服务研究，利用WebGL、HTML5、Ajax、Web Service、GPU并行计算等技术手段，通过探讨溃坝洪水时空模型计算优化、网络三维可视化模拟与动态交互分析等关键技术，研发了原型系统并进行应用实验。实验结果表明，本文研究成果有助于实现时空过程信息的网络发布，在线影响分析及动态可视化服务，可为分布式网络环境下溃坝洪水等时空过程信息管理和应急决策提供科学依据和技术支持。

英文摘要：

Many complicated geographic phenomena such as dam-break flood were complex giant systems, and the dynamic development of their spatiotemporal process is more important than the final formation of the spatial pattern. Therefore, the simulation and analysis of spatiotemporal process and the disaster forecasting warning have become a hot research direction in the field of geographic information science. Meanwhile, with the popularization of Web-GIS technology, the integration of real-time content on web and the visual expression of geographic spatiotemporal process simulation are also demanded by the public urgently. With the rapid development of web service, network technology and their applications in GIS, it is possible to realize the visualization analysis, share and manage distributed spatial data on different clients. After analyzing the latest Web-GIS technology, this paper focuses on the simulation and visual analysis of dam-break flooding spatiotemporal process in the network environment, which can be implemented by applying HTML5, WebGL, Ajax, Web Service technology and other technologies. Some key technologies, which include rapid computing of spatiotemporal process model, B/ S network architecture construction, three-dimensional scene rendering optimization, and dynamic interaction analysis, are also discussed in details. Finally, a prototype system was constructed and an experiment of the net- work simulation and visual analysis of a dam-break flood spatiotemporal process case was conducted on a study region. Experimental results showed that the methods addressed in this paper could successfully publish spatiotemporal process information, conduct online impact analysis and realize the three-dimensional visualization representation under the network environment, which could meet different requirements of browsing, querying and analyzing. It can provide an efficient support to dam-break management and emergency decision-making.