The positioning accuracy of gravity matching navigation is closely related to the resolu-tion of the gravity reference map,and the gravity reference map can be reconstructed by an inter-polation algorithm to improve the resolution.To verify whether the positioning accuracy is effec-tively improved by the refined gravity reference map reconstruction,the interpolation effects of multifaceted function interpolation,Shepard's interpolation and integrated Shepard's interpolation are compared and analysed.The three interpolation algorithms mentioned above are used to refine the gravity reference map,terrain contour matching(TERCOM)algorithm and iterated closest contour point(ICCP)algorithm are selected as the gravity matching algorithms.The gravity matching navigation accuracy of the interpolated gravity map is verified by off-line simulation of marine ship survey data.The experimental results show that after the resolution of the gravity map is increased from 1'×1'to 0.5'×0.5'by the three interpolation algorithms,the positioning accuracy of the TERCOM algorithm and the ICCP algorithm can be improved by about 30%and 20%,respectively.However,after the resolution is increased to 0.25'×0.25',the positioning ac-curacy is no longer significantly improved.In addition,although the interpolation effects of the three interpolation algorithms are different from each other,the navigation results are consistent,and the experimental results also show that the selection of the interpolation algorithms has little effect on the navigation results when the resolution of the reference map is only doubled.