World Stress Map Conference - International Lithosphere Program ...

3 rd World Stress Map Conference, 15.-17. October 2008 in Potsdam
Thursday, October 16 th
Oral Session III: Strain measurements and interpretation
The contemporary strain rate field of Fennoscandia
derived from BIFROST GPS
Hans-Georg Scherneck, Martin Lidberg, Jan M.Johansson, Rüdiger Haas, Glenn A. Milne
Chalmers University of Technology, Dept of Radio and Space Science
hgs@chalmers.se
We use up to 4800 days of BIFROST GPS in Fennoscandia and surroundings in order to
derive strain rates on a scale from 50 to 2000 km. Our investigations aim at discriminating
limitations of current models of Glacial Isostatic Adjustment (GIA). The BIFROST network
uses continuous satellite observations, mostly GPS, at 84 permanent GNSS stations, geodetic
quality sites in Fennsocandia and countries in the periphery. The time span of observations
covers 1993 to 2007. Data processing estimates daily positions and ancillary parameters, and
the solutions are constrained by and mapped into the international terrestrial reference frame.
From the data base of daily ositions, uncertainties and covariances, motion parameters for
each station are estimated with uncertainties on the order of 0.2 mm/yr in the vertical and less
than 0.1 mm/yr in the horizontal. We can demonstrate that the observations can be matched
with high accuracy using global models of the GIA process in Fennoscandia, models that
employ recent ice load histories and account for the dynamic evolution of the sea level in
response to the local and global GIA. We ask the question, to what extent the residual
deformation could be explained by more detailed models of the ice sheet history, by lateral
heterogeneities in the lithosphere/upper mantle structure and rheology, and by other (non-
GIA) processes, such as contemporary tectonics? We employ two different tools in order to
estimate maps of strain rates in Fennoscandia from the GPS results (i) minimising a strain
energy function using polynomials as base functions; (ii) employing a least-squares
collocation method. The strain rate calculation is considered as an efficient amplifier of
regional features of the displacement field. Our results indicate that shear strain is more
difficult to determine than areal strain. The former's error limit is at 8 to 10 nano/year while
observed shear strain in our analysis peaks at about 10 nano/yr. Since the GIA model predicts
only small levels of pure shear and no vorticity, we would expect non-GIA motion
components to be detectable notably in shear.
61