Digital Terrain Analysis Based On DEM

In recent years, Digital Elevation Model (DEM) as a digital representation method for topography has achieved a series of breakthrough in data capture, storage and management with the fast development of earth observation and computer technology. DEM has been one of basic contents of National Special Data Infrastructure(NSDI) in various countries and integrated by Numeric Spatial Data Frame(NGDF) to produce in large scale. Along with DEM being thoroughly developed with various resolution and large scale at national and provincial level, the acquisition possibility of all kinds of resolution and DEM with different precisions has been improved obviously. For this reason, various methods and technologies of terrain analysis need to be researched urgently for digging and using the abundant information contained in DEM to support decision making for other professional application fields.Terrain analysis consists of two kinds, namely the computation of basic terrain factor and the analysis of complex terrain factor. The computation of basic terrain factor mainly includes slope, aspect and so on. The analysis of complex terrain factor mainly includes visibility anlaysis, the extraction of topographic patterns and drainage network, etc. In this dissertation, the two sides of basic terrain factor and complex terrain factor are investigated and implemented based on the basic concept of digital terrain analysis introduced.With the basic terrain factor in this dissertation, slope, aspect, project distance, surface distance, project area, surface area, volume and slope analysis are discussed, and the common algorithms of these factors are given too. These factors are implemented in GeoGlobe 2.0. GeoGlobe 2.0 is a three-dimensional visualization software of the global and it is the first domestic software similar with Google Earth. The implement contains two parts, client and server, which communicate with other by HTTP(Hypertext Transfer Protocol).Intervisibility between two points is discussed and the computation of the intervisibility is described for the visibility analysis of complex terrain factor. The recurrence of the intervisibility is confirmed through the logical inference. Based on the above discussion, the common algorithms for viewshed computing are introduced, of which the advantages and disadvantages are analysised. Aiming at DEM database, a new improved Xdraw algorithm is put forward for viewshed analysis, then the accuracy and efficiency of the algorithm is shown from experimental comparison. For viewshed analysis based on DEM database, the algorithm having a good experimental effect reduces obviously the frequency of access to database with lower time complexity.Current algorithms are discussed concerning with feature extraction of complex terrain factor. For the hydrological flow modeling, depression identification and filling, flow direction determination over flat area are discussed in particular, and the method of flow direction determination over flat area is improved which efficiently reduce the flow parallel and denticle phenomenon. The algorithm of hydrological flow modeling is implemented. Simultaneausly, the deficiency and direction to contiue investigation of the algorithm are also pointed out.