在大数据背景下，人类对数据的利用由中尺度空间的建筑等延伸至裸眼无法直接完全观测到的宏观地球和微观分子尺度。对不同尺度数据的联合利用导致尺度上由宏观跨跃到微观的动态过渡状态的数据需求。对这种尺度宽广、数据量庞大的过渡数据进行组织和索引成为瓶颈。本课题拟从数据组织结构出发，研究基于Hash地理编码的全空间系统中多粒度时空对象数据模型。该模型对全球表面（10^7米）剖分至分子尺度10^-10米，依据空间位置构建基于多级明文Hash的地理编码对数据和存储节点进行索引。并组织动态树存储兴趣区域内邻居节点编码，分别对数据和节点异或运算比较逻辑距离，以获取各节点相对位置。该模型建立基于空间划分与地理编码索引的网络，将虚拟世界为代表的数据利用场景、现实地理空间和数据存储网络的位置映射，支持不同尺度间无缝过渡的数据获取方式。

In the context of big data, the requirement of data would extend from middle resolution space to macro comic resolution or molecular resolution that can not be directly observed by human eyes. The spatial data requirement would be changed dynamically from macro resolution to micro resolution because of the real-time data utilization jointly in different resolution. It is a bottleneck problem to dispatch mass of data with large resolution, and the smooth transition of data acquirement is also a challenge. The research topic is to invent a hash geocoding based data model of multi-granularity spatio-temporal object for pan-spatial information system. The model would subdivide global surface mesh(10^7 meter) into 10^-10 meter which is the molecular radius, and indexed by geographic code based on multistage Hash algorithm. Construct Hash Tree neighbor geographic code in interesting storage area, and compute the relative position among neighbors by XOR algorithm. Construct the data network based on space division and geo-code index and map the position of the models in real world and virtual world with the position in data network which could realize the real-time spatial data acquisition in the way of seamless transition between two diffrenet extreme resolutions.