Large Scale Drainage Network Extraction And River Structure Rules

As material and energy transport corridors, river systems play an important role in meteorological, hydrological, ecological and geomorphological relevant researches. The basic attributes, structure characteristics and geometrical parameters of a river system directly affect hydrological and ecological processes.As resolution increasing of raster DEM(Digital Elevation Model) data, traditional drainage network extraction algorithms have disadvantages in computation efficiency, accuracy and completeness. Moreover, approaches on river structures have limitations such as the precision and scale of available river networks and the representativeness of river network samples. Therefore, it is desired to undertake approaches on river network extraction methods for large-scale river basins with high-resolution DEM data, and explorations of river structures and corresponding geometric characteristics.To solve the shortages of existing drainage network extraction methods in computation efficiency and accuracy resulting from high-resolution DEM data, in this study, a time-efficient and comprehensive method has been proposed for extracting drainage networks from high-resolution DEM data in large-scale river basins, through introducing a size-balanced binary search tree. The developed method has been applied to real river basins to verify its advantages in time-efficiency, accuracy and completeness of obtained river networks. To effectively manage the extracted high resolution river networks and corresponding massive data, based on the stream ordering method, a hierarchical pyramid method has been proposed through gener ating high level river networks from that under lower level with proposed stream merging algorithm. A river network codification method has also been established to identify streams in river basins from each other. Finally, through applying the above devel oped methods to most river basins on the earth, a high-resolution digital river network dataset, the Hydro30 dataset, has been obtained.To characterize complex structure of a river system, based on the proposed hierarchical pyramid method, a hierarchical description method has been proposed through extracting topologic information between river networks under consecutive levels, defining basic units of the river network structure and building a structure unit pyramid of hierarchical river networks. To reve al the topologic and geometric rules of river network structures, 127 large river network samples have been selected from the obtained Hydro30 dataset. Through analyzing six structure parameters from traditional description method, and the distribution characteristics of eight parameters of the structure unit pyramid from proposed hierarchical description method, a general statistical phenomenon that the selected rivers have consistent structure characteristics have been found. The results indicate that riv er structures follow specific topologic and geometric rules which are represented by relevant factors such as the length, area, slope and angle.The developed river network extraction algorithm, massive river network data management method, corresponding global high-resolution river networks(the Hydro30 dataset), and the determined topologic and geometric rules of river structures, provide not only basic data support for hydrological process simulation, water resources management and river regulation resea rches, but also important scientific basis for further discovery of inherent essence of natural river structures.