High Accuracy Surface Modeling （ HASM） method

based on the fundamental theorem of surfaces

provides a solution to the error problem that long troubled geographic information science. Although HASM has a sound theoretical foundation

the low computing efficiency arising from solving large scale linear systems

poses a severe limitation on its wide-spread applications in construction of large scale and high resolution surfaces. In order to achieve a high speed

we developed an algorithm based on two dimensional Double Successive Projection Method （ HASM-DSPM）

which significantly improves the performance by adopting strategies of project space adjustment and two-step correction. HASM-DSPM is examined by both numerical experiment and real world experiment

the results of which both illustrate that HASM can surpass Gauss-Seidel （ GS ） and Modified GS （ MGS ） in convergence pace and computational speed. Furthermore

compared with other classical surface modeling methods

our method has obviously higher accuracy.