Task 1: Analysis of SAR images from the ERS-1 and ERS-2 satellites

Start: March 1996 (month 1)
End: September 1997 (month 19)
Responsible partner: CNRS.DTP
Cooperative partners: NVI, UICE.DG

Kurt L. Feigl, Freysteinn Sigmundsson, Helene Vadon and Didier Massonnet

The objective of this work (according to the revised work description) was to calculate and interpret Synthetic Aperture Radar (SAR) interferograms to capture interseismic deformation, spanning at least one year.

This goal has been reached and even exceeded. We have studied two areas in Iceland using radar interferometry, the Reykjanes Peninsula oblique plate boundary in SW-Iceland and the Krafla volcanic system in Iceland. In both areas we have captured interseismic deformation spanning up to three years, thereby validating the use of radar interferometry for the monitoring of crustal deformation in the seismic and volcanic zones in Iceland. The results help to understand the mechanism of oblique rifting, the behaviour of volcanic zones in post-eruptive periods, and the generation of earthquakes. The work was conducted in collaboration with Didier Massonnet and Helene Vadon at CNES, Toulouse. Initially a number of interferograms were calculated at CNES for each of the two study areas. These interferograms have then been successfully modelled, interpreted, and published in reviewed scientific journals of international repute.

In the first study area, the Reykjanes Peninsula in SW-Iceland, we have used satellite radar interferometry to map the satellite-view component of a plate-boundary velocity field, as well as volcano deformation. The area is the direct onland structural continuation of the submarine Mid-Atlantic Ridge. Oblique spreading between the North American and Eurasian plates of 1.9 cm/year occurs there, causing both shearing and extension across the plate boundary. Using ERS-1 images from the 1992-1995 period we have formed interferograms, spanning up to 3.12 years. The fringes in these interference patterns are clear, indicating that the dielectric properties of the ground surface in this area remain unchanged. Moreover, these fringes measure crustal deformation. The most obvious deformation is secular deflation of the Reykjanes central volcano, averaging to 15 mm/year, probably caused by compaction of a geothermal reservoir in response to its utilization by a power plant. The deflation we infer is in good agreement with levelling data. This gives confidence in the interpretation of more subtle deformation signal in the interferograms, fringes aligned in the direction of the plate boundary caused by plate boundary deformation. Relying partly on geologic evidence, we assume the shape of the horizontal and vertical crustal velocity field. We estimate best-fit model parameters by maximizing the global coherence of the residual interferograms, the difference between observed and model interferograms. The data constrain the locking depth of the plate boundary to be about 5 km. Below that level the plate movements are accommodated by continuous ductile deformation, not fully balanced by inflow of magma from depth, causing about 6.5 mm/year subsidence of the plate boundary. Previous regional geodetic data agrees with this interpretation.

For the other study area, the Krafla spreading segment in North Iceland, we have used radar interferometry to map deformation in the 1992-1995 period, related to readjustement of the spreading segment to crustal rifting. Three images acquired by the ERS satellites were used to produce 3 interferograms. Time-average deformation in the 1992-1995 period amounts to 24 mm/year subsidence above a shallow magma chamber at Krafla, superimposed on 7 mm/year along-axis subsidence of the spreading segment. The rate of subsidence decays with time, it appears two times lower in 1993-1995 than in 1992-1993. Several process may contribute to the subsidence, all which relate to ongoing post-rifting adjustment of the spreading segment after a rifting episode in 1975-1984. Subsidence above the magma chamber may relate to magma solidification and cooling; solidification of 10⁶-10⁷ m³/year of magma is required to fully explain it. Processes likely to contribute to subsidence along the Krafla rift are ductile flow of material from the rift axis and cooling. Cooling by 0.5$^{\circ }$C per year, the measured cooling of groundwater at the rift, of 1.5 km wide and 3 km high zone along the rift axis could explain the subsidence. The time decaying subsidence may also reflect time decaying ductile flow of material from the rift axis. The relative contributions of cooling and ductile flow to the subsidence is difficult to assess, but both processes indicate the readjustment of the spreading segment to crustal rifting. The current low rates of deformation at Krafla suggest the readjustment is in a final stage, a decade after the end of the 1975-1984 rifting episode.