Halfdan P. Kierulf

A GPS velocity field for Fennoscandia and a consistent comparison to glacial isostatic adjustment models

In Fennoscandia, the process of Glacial Isostatic Adjustment (GIA) drives ongoing crustal deformation. Crustal velocities from GPS observations have proved to be a useful tool in constraining GIA models. However, reference frame uncertainties, plate tectonics, intraplate deformations as well as other geophysical processes contaminate the results. Former studies have shown that different International Terrestrial Reference Frames have large discrepancies, especially in the vertical component, which hamper geophysical interpretation. We present new velocity estimates for the Fennoscandian and North European GPS network. Our GPS velocity field is directly realized in a GIA reference frame. Using this method (named the GIA frame approach) we are able to constrain GIA models with minimal influence of errors in the reference frame or biasing signals from plate tectonics. The drawbacks are more degrees of freedom that might mask real but unmodeled signals. Monte Carlo tests suggest that our approach is robust at the 97% level in terms of correctly separating different models of ice history but, depending on deformation patterns, the identified Earth model may be slightly biased in up to 39% of cases. We compare our results to different one- and three-dimensional GIA models employing different global ice-load histories. The GIA models generally provide good fit to the data but there are still significant discrepancies in some areas. We suggest that these differences are mainly related to inaccuracies in the ice models and/or lateral inhomogeneities in the Earth structure under Fennoscandia. Thus, GIA models still need to be improved, but the GIA frame approach provides a base for further improvements.

A continuous velocity field for Norway

In Norway, as in the rest of Fennoscandia, the process of Glacial Isostatic Adjustment causes ongoing crustal deformation. The vertical and horizontal movements of the Earth can be measured to a high degree of precision using GNSS. The Norwegian GNSS network has gradually been established since the early 1990s and today contains approximately 140 stations. The stations are established both for navigation purposes and for studies of geophysical processes. Only a few of these stations have been analyzed previously. We present new velocity estimates for the Norwegian GNSS network using the processing package GAMIT. We examine the relation between time-series length and precision. With approximately 3.5 years of data, we are able to reproduce the secular vertical rate with a precision of 0.5 mm/year. To establish a continuous crustal velocity field in areas where we have no GNSS receivers or the observation period is too short to obtain reliable results, either interpolation or modeling is required. We experiment with both approaches in this analysis by using (i) a statistical interpolation method called Kriging and (ii) a GIA forward model. In addition, we examine how our vertical velocity field solution is affected by the inclusion of data from repeated leveling. Results from our geophysical model give better estimates on the edge of the network, but inside the network the statistical interpolation method performs better. In general, we find that if we have less than 3.5 years of data for a GNSS station, the interpolated value is better than the velocity estimate based on a single time-series.