Quasigeoid computation in the Nordic and Baltic region

The Nordic and Baltic Sea region is topographically varied, making gravity potential field modelling in the region a challenge. With new gravity data in and around the Baltic Sea, this study investigates improvements in the accuracy of quasigeoid determination in the region. The quasigeoid determination method is the remove-compute-restore using a spherical fast Fourier transform to solve Molodensky's integral, using a Wong-Gore kernel modification. Different kernel modification ranges of the Stokes' kernel are tested. The gravimetric quasigeoids are compared to global navigation satellite systems (GNSS)-levelling datasets in the region to assess the accuracy of the models. The best fitting model has a normalised root mean square error (RMSE) of the residuals to GNSS-levelling height anomalies for a subregion of below 2 cm, while the normalised RMSE to GNSS-levelling height anomalies in the whole region is 2.8 cm, matching the official regional gravimetric Nordic geodetic commission quasigeoid from 2015. A simple planar tilt correction of the residuals improves the fit in Denmark and Latvia. The correction indicates systematic tilts in the national GNSS-levelling networks of said countries. The investigation of different kernel modification ranges shows that the long transition intervals are less sensitive to the choice of bounding degrees, compared to short intervals, and more robustly give a better fit to the GNSS-levelling data in the region.