Geological tetrahedral model-oriented hybrid spatial indexing structure based on Octree and 3D R*-tree

Abstract

The lack of efficient access and query method remains a barrier for three-dimensional (3D) geological tetrahedral models (GTMs) to support high-performance 3D spatial analysis and calculation. To organize and manage GTMs efficiently, this paper proposes a GTM-oriented hybrid spatial indexing structure, named 3DOR*-tree, which can fully exploit the advantages of the fast spatial partition of Octree and the efficient spatial query of 3D R*-tree. GTM-oriented data structures are designed on the basis of 3DOR*-tree by setting appropriate thresholds for construction. Geological spatial data query based on 3DOR*-tree can then be implemented. Through test verification and comparative analysis, the accuracy and efficiency of the proposed spatial indexing structure are verified. The node-splitting thresholds of the 3DOR*-tree indexing structure are discussed and analyzed. The 3DOR*-tree can implement efficient construction and query simultaneously. Results corroborate that the node-splitting threshold of Octree is recommend to be set as 500 tetrahedrons, and that the node-splitting thresholds of 3D R*-tree are recommended to be set as 128 and 256 individuals (minimum and maximum tetrahedrons, respectively). This study not only provides technical support for the storage, management, and analysis of GTMs but also provides reference for the theoretical research of a hybrid spatial indexing method.