Abstract: Modelling the spatial correlation error Accurately is the key of network RTK technology. Quantitative analysis of the interpolation models’ robustness is very important for the selection and optimization of network RTK interpolation models. The robustness of the interpolation model reflectingIn this paper, the existing interpolation models were studied, based on the law of error propagation, the concept of error impact factor is proposed to evaluate the robustness of the interpolation models. The theoretical and experimental methods were used to explore the spatial distribution characteristics of interpolation models’ robustness. The experimental results indicate that there is a great difference in the robustness of different interpolation models in spatial distribution, and inside the triangular cell, different interpolation models both meet the robustness requirements.