The application of gridding method for digital elevation model

In this study, it is searched to find the optimum approximation method to reproduce DEM inthe regions lacking of DEM datum by using grid placing algorithms. Moreover, after findingthe optimum approximation method, it is tried to be determined the effect of the choice of thegrid spacing, the importance of amount of known points left at the insufficient datum regionand the effect of area magnitude without any DEM data to the accuracy. This study consists of5 main chapters.In the first chapter an overview of the research is presented. The goal of the research andapplications related on the study are summarized as well.In the second chapter, definition of DEM, its importance, quality, advantages, disadvantages,accuracy and the way of storage are given. In addition, the separation of DEM into subgroupsin the area is tried to be explained.In the third chapter, grid placing algorithms used for producing SYM is explained boththeoretically and visually.In the fourth chapter, the accuracy of approximation method is examined by applying of 5various grid placing algorithms.The first application is about which algorithm selected as approximation method assures theaccuracy of the reproduced DEM and searching the most useful method for each selectionarea.Second application is about the issue of what will be the grid placing space of the algorithmthat is chosen as optimum. The DEM data produced with each 20 meters space from aerialphotographs assumed as accurate and reliable are compared with the SYM data reproducedfrom changed grid placing space.The third application is about the issue of determination of the amount of known points at theselected field and at how many points, how much accuracy can be gained.In the fourth application, the importance of the area in determination of the selected area ismentioned. The change in accuracy criteria by extension of the field is searched nearby.Finally, the results of this study and possible suggestions related to them are mentioned in thelast part of this research.With this study the usuability of the ways of approximation methods esspecially kriging,radial bases and inverse distance is tried to be proved. The practices of these methods areproved by the statistic values.In the study Golden Surfer 8.0 Erdas Imagine 8.7 and kinds of interface programs are used.Besides there are points which are sellected to define the flat and mountanıous areas calledsellection zone which are accepted without any data. During the point sellection activity theinterface programs which are built up to omit the unnecessary datum such as cloudy areas, areused. The most time taking part of the thesis is this point sellection activity. Erdas Imagine 8.7is used in the processes such as to view the images of the satellite, to determine the sellectionregion, to convert the DEM datum to excel formation and etc. Golden Surfer is also used toasa tool to reproduce the lacking DEM datum sorted by interface programs which will be usedfor the study by various approximation methods statistically. Golden Surfer is choosen that?swhy it includes appoximation methods much more. For example Erdas Imagine or Arcviewprograms just only consider kriging or inverse distance. And this could only be done bypreparing special scripts.As a result the usuability of the approximate methods which has applied in mine geodesicdepartment, geodesy studies, climate changes, figuring out the pollution changes under thesome kind of water sources such as lakes sea etc. and soil erosion of snoping fields and alsoon the studies of fotogrametic studies over the DEM datum is being proved.Kriging method is found the most suitable approximate method with ±0.78 m on the flat areasat 1457 points, with ±12.5 m on mountanious areas at 5945 points and then the radial basesmethod was found out. As the amount of the known points which are left on the areas calledas sellection zone and has no itendified DEM data, increases the truth algorithm. It descreaseswhen the zone gets larger. The main point is that when the dispersion of the height whichbelongs to the area is homogenious it increases the turth every time as the most effectivepoint. The reliability of the approximate method increases when the value difference betweenthe max and min heights is little. For example while this amount is little on flat areas it getsmuch more on mountanious areas. Up to this information with the approximate algorithm thedata which is carried out on flat areas are close to the original one. New reliable datum can befound out any time when the DEM data is lacking or wrong by profitting the approximatemethods which are used in the thesis. And this brings saves of money and time nearby. Forexample if there is a cloud over the satellite view there is no need to wait for the satelliterepassing over the same area to get DEM of it. By retaking the view the save of money isgained. There isn?t any firm or association in our country that uses the kriging or radial basesmethods to get DEM. The most suitable approximate algorithm with these results and thepoints that effects them, the aim is to guide the military and civilian areas scholarly