项目名称: 时间密集型空间大数据实时计算关键方法研究——以精细气象预报产品快速生成为例
项目编号: No.41301438
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 天文学、地球科学
项目作者: 董文
作者单位: 中国科学院遥感与数字地球研究所
项目金额: 25万元
中文摘要: 针对大数据汹涌背景下空间大数据的快速处理及知识发现问题,本项目主要以时间密集型的非结构化空间数据为主要研究对象,从多源异构空间数据在不同时间尺度上的置信度和生命周期差异角度出发,提出一种时间尺度分层模型,基于这种时间尺度分层理论,研究具有时间序列的空间数据分层组织管理模型,并在此基础上进一步开展时间密集型空间大数据的实时接入方法和精细空间信息产品生成方法研究,发展云IS平台支持下的时间密集型空间大数据实时计算技术,建立时间密集型空间大数据快速处理的方法和技术体系,并结合气象降水预报及洪水预警需求,开展技术验证实验与应用研究。通过对相关理论方法的研究及技术实现,一方面显著提升我国新一代GIS的核心技术水平和国际竞争能力,占据新一代GIS发展的技术前沿,另一方面支撑国家、区域及地方不同层次的气象灾害预报预警及应急决策应用。
中文关键词: 云计算;大数据;多源数据集成;传感网数据;决策器模型
英文摘要: Rapid processing and knownledge discovery are urgently and essentially needed under the background of spatial big-data explosion. From the perspective of discrepancies in confidence level and in life-cycle of the hetorogeniously multi-sources spatial data for different time scales, this project propose a hierarchical timescale framwork with mainly focussing on the time-intensive unstructured spatial data. Under the framwork of this hierarchical theory, a layered data management model based on time-series spatial data is developed, and the real-time access of the time-intensive unstructured big data and the fast generation of fine-grid spatial products are deeply explored. Further, the real-time computing of time-intensive big data based on the Cloud GIS platform and fast data processing technology system are established. Finally, the validation and application of these technologies are conducted in combination with the need of meteorological precipitation forecast and flood warning.By acomplishment of the relevant theories and technologie, this study ,on one hand ,significantly improve China's new generation of GIS technology level and international competitiveness; on the other hand, it can support the national, regional and local meteorological disaster forecasting, early warning and emergency decisi
英文关键词: cloud computation;big data;multi-sourcespatial data integration;sensor network data;Decision maker model