Abstract In this paper, the similarity transformation method and the strong constraint method are used in processing the extended 260 CGCS2000 skeleton points with the GAMIT/GLOBK software. Coordinates of stations in the ITRF2008 are obtained through adjustment calculation adopting three schemes. The comparison result of the three schemes shows that different constraints on the original data will result in different positioning precision, and that inappropriate constraints may lead to network deformation and distortion with the strong constraint method. The similarity transformation method is more reliable and can avoid the above-mentioned problems. Its average mean square error is less than 1 mm.
