The article presents a methodology for transitioning to the Amsterdam height system based on utilizing global and regional geoid/quasi-geoid models. The study was conducted for the Polish-Ukrainian cross-border sector and expanded to the entire territory of Poland and Ukraine. The input data comprised two regional and five global geoid/quasi-geoid models. The initial data analysis was conducted for all models relative to GNSS/leveling data in the Baltic height system. The secondary analysis was performed relative to the combined PL-quasi-geoid2021 model and the gravimetric EGG2015 model. Based on the analysis results, a methodology for optimizing heights between regional and global geoid/quasi-geoid models was developed, including the following stages: calculation of conditional global and regional geoid/quasi-geoid heights, calculation of approximately predicted height differences between the conditional regional and global geoid/quasi-geoid, implementation of the refinement (correction) of approximate heights at the regional model, optimization of approximate heights at the regional model, and calculation of a regional combined model in the Amsterdam height system. The developed methodology enables the integration of regional and global geoid/quasi-geoid models into the Amsterdam height system with an accuracy of 1–2 cm by optimizing their heights. The advantage of this methodology is that it requires only a minimal amount of GNSS/leveling data to establish connections between different height systems.

At present, one of the methods used to determine the height of points on the Earth's surface is Global Navigation Satellite System (GNSS) leveling. It is possible to determine the orthometric or normal height by this method only if there is a geoid or quasi-geoid height model available. This paper proposes the methodology for local correction of the heights of high-order global geoid models such as EGM08, EIGEN-6C4, GECO, and XGM2019e_2159. This methodology was tested in different areas of the research field, covering various relief forms. The dependence of the change in corrected height accuracy on the input data was analyzed, and the correction was also conducted for model heights in three tidal systems: “tide free”, “mean tide”, and “zero tide”. The results show that the heights of EIGEN-6C4 model can be corrected with an accuracy of up to 1 cm for flat and foothill terrains with the dimensionality of. The EGM08 model presents an almost identical result. The EIGEN-6C4 model is best suited for mountainous relief and provides an accuracy of 1.5 cm on the area. The height correction accuracy of GECO and XGM2019e_2159 models is slightly poor, which has fuzziness in terms of numerical fluctuation.