The Statistical Analysis of the Eigenspace Components of the Strain Rate Tensor Derived from FinnRef GPS Measurements (1997–2004) in Fennoscandia

Abstract

In the deformation analysis in geosciences (geodesy, geophysics and geology), we are often confronted with the problem of a two-dimensional (or planar and horizontal), symmetric rank-two deformation tensor. Its eigenspace components (principal components, principal direction) play an important role in interpreting the phenomena like earthquakes (seismic deformations), plate motions and plate deformations among others. With the new space geodetic techniques, such as VLBI, SLR, DORIS and especially GPS, positions and change rates of network stations can be accurately determined from the regular measurement campaign, which is acknowledged as an accurate and reliable source of information in Earth deformation studies. This fact suggests that the components of deformation measures (such as the stress or strain tensor, etc.) can be estimated from the highly accurate geodetic observations and analyzed by means of the proper statistical testing procedures. We begin with discussion of the geodynamic setting of the selected investigated regions: Fennoscandia. Then the regular GPS observations in Finnish permanent GPS network (FinnRef) and the related data preparation are introduced. Thirdly the methods of derivation the two-dimensional geodetic strain rates tensor from the surface residual velocities and the newly developed estimates BLUUE of the eigenspace elements and BIQUUE of its variance-covariance matrix are reviewed. In the case study both BLUUE and BIQUUE models are applied to the eigenspace components of two-dimensional strain rate tensor observations in Fennoscandia, which are derived from 1997 to 2004 annual station positions and velocities of FinnRef. Further detailed analysis of the results is also performed with respect to geodynamical and statistical aspects.