Abstract:In view of describing the territory of China as a whole from a statistical perspective, this paper gives an estimate of the global spatial characteristics of the territory of China using global spatial statistics (or statistics of spatial distribution). A design-and pattern-based method of spatial sampling, or the global spatial random sampling, is proposed, and applied to the China's Digital Elevation Model (DEM, 1 : 4 000 000). The parameters of the spatial pattern, including the center of gravity, orientation, geographical range and spatial intensity, are estimated for the DEM, which is decomposed into a terrain surface and a geometric outline. The results demonstrate that both the surface and the outline differ remarkably in the center of gravity, orientation, geographical scope and spatial intensity, revealing the complexity of the territory of China on the global surface; the small margins of errors show the efficiency and effectiveness of the proposed pattern- and design-based approach of spatial sampling.