Comparative Analysis And Collaborative Application Of NoSQL And Relational Spatial Database

In recent years, with the continuous development of remote sensing technology and geographic information system (GIS) technology, GIS was used more and more widely. The amount of spatial data in the GIS grows rapidly. More and more big data and unstructured data spring up.This presents a huge challenge to traditional spatial data management always realized with the relational database or object-relational database.With the advent of the Web2.0era, to meet the application requirements of the explosive growth of the Internet, NoSQL database management technology arose. NoSQL database has advantages in terms of storage, query and management of big data and unstructured data. In this paper, comparative study was taken between relational and NoSQL spatial database.In this paper, firstly the architecture, storage mechanism and mode of operation of MongoDB, ArcSDE (Oracle) and PostGIS were analyzed. Secondly qualitative and quantitative evaluation factors were selected, evaluation system was established and comparision experiment was taken. Thirdly, based on the result of comparion. the characteristics of the NoSQL and relational data spatial database in spatial data management were summarized and the NoSQL and relational database collaborative applications schema was design. Fourthly, with the OGC Simple Feature Model and object-oriented and component technology, the key technology of the schema was relized. The collaborative applications schema was used in earthquake rescue data management system and evaluated the schema applicable value.In this paper, the main conclusions and results are as follows:1. Through the establishment of quantitative and qualitative evaluation system, a comparative study on MongoDB, ArcSDE and PostGIS was taken. And drawn the conclusion that the MongoDB consumes least time when stored the same size vector data than the ArcSDE and PostGIS. However stored the same size raster data the ArcSDE performs fastest and consumes less storagement capacity than other two databases. In aspcts of spatial query, the MongoDB performs more efficient than the other two, the greater the query range, the more efficient it will be.2. Based on the results of the comparative analysis, the collaborative application schema was proposed.The MongoDB used to mange the vector spatial data and spatial metadata whileArcSDE used to manage the raster data. And the key technologies of collaborative schema were implemented. The spatial data in NoSQL, migrate spatial data between NoSQL and relational database and mange the both two database with the unified interface were implemented3. According to the needs of earthquake rescue, data management.The NoSQL databases and relational database collaborative appilcation schema was used in the earthquake rescue data management system. Based on the MongoDB and ArcSDE collaborative application shcema designed and implemented the earthquake data management system. The evaluation of the system indicated that the NoSQL and relational database collaborative application improved earthquake rescue data storage and query efficiency and reduced data storage disk consumption, conviniented the database update and expansion. It proved the collaborative application of NoSQL and relational database better meet the earthquake rescue data management needs.