Distributed Multi-temporal Geoinformation Service Aggregation Based On A Compound Pyramid Model

This paper analyzes and applies the foundational theories and concepts behind geo-information services. Focusing on the inherent features of the distributed, multi-sourced, multi-scalar, multi-temporal and heterogeneous architectures of geo-information services as employed to construct the MapWorld project, and in response to the actual application demands for geo-information service aggregation, and a service sharing model with high performance and high concurrence, following the OGC standards, this paper proposes a theory and methodology for distributed multi-temporal geo-information service aggregation, providing theoretical and technological support to facilitate the implementation of large-scale geo-information public service projects.The key research findings include:(1)A compound pyramid model was proposed and patented. This model is described in universal XML format, and by nesting the basic meta-information of several pyramid models, realizes descriptive content within a compound pyramid structure containing multiple pyramid schemes, making one pyramid structure suitable for multi-scalar and multi-sourced tiled dataset generation.(2) A methodology for distributed multi-temporal geo-information service aggregation was proposed, with two patents pending. Given multiple distributed web map tile services, the method logically organizes and aggregates them into an Aggregated Web Map Tile Service (AWMTS), through which a client program can rapidly locate and access the requested services to better realize one-stop service provision. With a multi-temporal tiled dataset design, the method unifies a tiled dataset model, allowing users to browse tiled datasets for any time and at any extent. Multi-temporal tiled datasets can be overlapped for visualization, providing a convenient way to compare and analyze temporal changes as shown by the tiled datasets from different time periods.(3) Based on the proposed compound pyramid model and methodology described in this paper, a prototype system was realized in a large-scale geo-information service platform.(4) These key technologies were tested during the construction of MapWorld, the national geo-information public service platform organized by the NASMG (National Administration of Surveying, Mapping and Geoinformation), and is proving to be effective with tremendous social and economic benefits.The key innovations include:(1) On the basis of a titled pyramid structure, a compound pyramid model was proposed. Based on this model, component interfaces and tiled map services were designed. The compound pyramid model enables a universal pyramid model description for tiled datasets generated from multi-scalar and multi-sourced geo-data. As compared to the traditional approaches using heterogeneous pyramid models for different kinds of scaled data with incompatible systems and from different organizations, the new model is better suited, especially in large-scale data situations, for geo-data updating (including adding new types of data), and does not require the design of an extra pyramid model, but only an update of the corresponding parameters in the universal compound pyramid model.(2) A methodology for distributed multi-temporal geo-information service aggregation was proposed and applied to realize a prototype system in the GeoGlobe environment. Based on a service proxy model, users access and request different kinds of geo-information services via a service proxy. The service proxy, in accordance with user demands, searches for the best service to return for a user call. The methodology supports aggregation of multi-temporal data services. In contrast to the traditional method using multiple datasets for each historical period, this paper proposes a new approach that uses just one dataset to describe multi-temporal datasets. Data for different historical periods are taken as subsets of this single unified dataset. When a data update is needed, a new subset is merely added to the unified dataset, accordingly. By adopting this approach, the issue arising from frequent changes in data services is resolved with improved efficiency for geo-information updates.The research findings detailed in this paper illustrate a feasible technical route towards large-scale multi-layered geo-information service provision.