The Study about the Algorithms of Automatic Drainage Extracting and Software Design Based on Triangulated Irregular Network

Abstract: Digital Elevation Model (DEM) is the fundamental and technical platform of hydrology research in building digital riverbasin. Basing on the basic data resource of DEM, various topographical characteristics associate with drainage area could be automatically derived such as slop, flow direction, drainage density, drainage structure and watershed shape etc., automatic terrain partition of drainage and calculational cell of distributed hydrological model could also be obtained form the result above. Drainage analysis based on DEM is the foreland and direction of hydrological simulation at digital ages.At the present time, most algorithms are designed for raster-DEM and have been widely used in GIS software and hydrology software. Triangulated Irregular Network (TIN) is also a main format of DEM. Comparing with grid based DEM, TIN has some advantages such as transformable resolution, high accuracy of terrain modeling etc, but the TIN was not get wide applications in hydrological model for its complexity in data structure. This paper presents an algorithm for extracting drainage network on a terrain model defined by TIN. In the paper, the watershed mode that was needed for drainage network extraction was discussed firstly. Then based on the watershed mode, the paper presented the relationship of water flow direction and the triangle side in TIN, analyzed the confluence character of the triangle side and the identified rules. Lastly, we designed the algorithms that can be used to extract drainage network and subcatchment from TIN. The experimental program is designed for validating the correctness and ATC (Asymptotic Time Complexity) of the algorithms, and the results of the research indicate that the algorithms are creditable…
Key words: DEM; TIN; drainage; drainage extracting; subcatchment extracting; hydrological analysis

This entry was posted in Master Thesis. Bookmark the permalink.