Research On The Method Of Statistical Analysis For National Geographic Condition Based On Cloud GIS

Statistical analysis for national geographic condition is the significant method to obtain the deep geographic condition information from the bulky data in China's first geographic conditions survey.The current period represents a momentous turning point in the process of economic and social structure upgrading of our state,and a crucial time for rapidly and increasingly complex changing in the national conditions,making the comprehensiveness and currency of results in statistical analysis vital for scientific decision-making and management planning,which become the soul and vitality of statistical analysis consequently.Based on the fact that there are some limitation in expansion of data storage,processing performance,disconnected systems and service in the current Statistical analysis work for national geographic condition,the research work of this dissertation is aimed to establish a elastic and extendable service system under the environment within cloud GIS,providing a flexible and efficient method for statistical analysis for national geographic condition,which combines the technology of web cloud service and high-performance computing.Specific study as follows:(1)The dissertation designs a extensible model of multi-level service system for statistical analysis for national geographic condition to solve the problem of isolated systems,redundant functions and single applications through on-demand invoking and the assembly of multi-granularity services.Variety of services and resources are both effectively integrated into the system to support diversified statistical analysis with high-reusability,high-flexibility and high-scalability services,which is able to break the barriers of data and information.(2)The dissertation constructs a distributed storage pattern based on HDFS and Hbase for massive and multi-source survey data to solve the bottleneck of storage of existing reletional database run on a single machine,realize centralized management and maintenance of census data and provice a efficient data response.(3)Based on the cloud environment of Hadoop,a parallel processing model of statictical analysis for geo-national condition is designed to break the bottleneck of single-machine process in comlex spatial analysis and statictics of large-scale features with the parallel algorithm of statistical analysis using MapReduce technology.The pattern of parallel processing achieves a quasi-real-time service response and efficient data processing.(4)A integrated service platform for statictical analysis for geo-national condition is established including massive census data's cloud storage,high-performance computing,and other business application relying on the multi-level service system,distributed database and parallel processing architecture.And then a client application instance is built to prove that the proposed scheme can meet the needs of the extensible storage and the efficient execution of statistical analysis tasks for bulky census data.