基于计算网格的DEM空间分析系统

计算网格技术可以解决DEM空间分析中大数据量和高密度计算对计算资源的需求。计算网格平台Alche-mi给出了模型系统的架构设计和层次设计,将系统划分为DEM分析服务器、计算网格节点和空间数据库服务器;通过集中式空间数据库服务器和基于数据分解的方法设计了系统的并行算法。实例结果表明,应用网格技术可以显著提高大计算量DEM模型分析的运算效率,具有重要理论和现实意义。

The DEM Analysis Tools Based on Computational Grids

Along with the multi-scale digital elevation model (DEM) accomplished in China,the DEM-based spatial analysis plays an increasingly important role in GIS research,and the DEM-based spatial analysis needs high resolution data,which means the increasing demand for density computing.Computational grids can solve the high density computing demand for DEM analysis that involved enormous amounts of data.We demonstrated the feasibility of DEM analysis tools and completed the prototype of the system based on computational grids and designed system architecture based on Alchemi,which is a computational grids framework.The DEM analysis tools can be divided into DEM analysis server,computational grids nodes and spatial database server in architecture.The DEM analysis tools can be implemented in API,grids,spatial analysis and application layers.The Parallel algorithm of the tools is based on the principle of procedure decomposition,functional decomposition and data decomposition.A separate spatial database server is designed as a spatial data server to avoid huge decline in efficiency caused by the mass data transfer in distributed computing grid.The efficiency of computational grids is tested by slope factor calculation based on the prototype system of DEM analysis tools.Experiments show that the efficiency of DEM analysis for big dataset can be significantly enhanced by computational grids,but for small dataset,the efficiency is not significant.The results confirm the feasibility of the application of computational grids to DEM spatial analysis for enormous amounts of data.