Interseismic Deformation

Earthquakes cannot occur without the build up of elastic strain. It is in mapping the accumulation of this interseismic strain that InSAR offers the most potential as a medium-range forecast tool. Note that I carefully avoided the term prediction, as this implies something like a 2-day earthquake warning. An earthquake forecast would give the likelihood of an earthquake occurring over a certain time period (e.g. "There is a 62% chance of strong shaking in greater Istanbul in the next 30 years"). These medium-range forecasts are vital because they enable civil defence agencies to prepare communities through education, rebuilding and retrofitting programs.

Measuring the build up of elastic strain between earthquakes (inter-seismic strain) using InSAR is not straightforward. The strain rates are extremely small — the North Anatolian Fault, for example, moves at around 24mm/yr horizontally and it therefore takes around three years to create a single interference fringe. Interseismic fault creep, where slip continues to the surface, produces a discontinuity that is relatively straightforward to observe in interferograms [e.g. Rosen et al. (1998)]. In contrast, interseismic deformation associated with faults that are locked at the surface is typically distributed on a length scale of 30-150 km and therefore much harder to distinguish from atmospheric and orbital errors.

In an ideal world, we would look at interferograms spanning a very long time interval, but, in areas such as Turkey, interferograms with time intervals of larger than 2 years are generally incoherent. We are therefore restricted to shorter-period interferograms, but these contain such a small deformation signal that they tend to be swamped by noise. To overcome this problem requires the use of multiple interferograms to amplify the tectonic signal and reduce the noise. By summing several interferograms, I was able, with colleagues, to extract the pattern of strain accumulation across the North Anatolian Fault [Fig. 6; Wright (2000); Wright et al. (2001a)]. The stacked interferogram is effectively an image of the gradual build up of elastic energy in a ca. 70km wide zone across the North Anatolian Fault. This energy will eventually be released in an earthquake. The image is also a direct observation of plate tectonics in action, revealing the relative motion of Anatolia with respect to the Eurasian Plate.

Elsewhere, Peltzer et al. [2001] have used similar methods to measure the deformation of the San Andreas Fault Zone in southern California. My colleagues and I have also used InSAR to show that the present-day rate of strain accumulation on the major faults in western Tibet is lower than expected from geological observations [Wright et al. (2004)].

With current satellites, it is only possible to use InSAR to measure strain accumulation if ground conditions are optimal. In particular, vegetation cover must be relatively low. With future satellite technology, it will be possible to build-up a time-varying map of strain across most of the

Fig. 6 (a) Topographic and tectonic map of the eastern end of the North Anatolian Fault. The coloured area is an elevation model calculated from a 1-day interferogram: the North Anatolian Fault (dashed red line) can clearly be seen cutting through the landscape. Arrows are GPS-determined velocities relative to the Eurasian plate (McClusky et al., 2000); (b) Stacked inter-seismic interferogram, converted to a yearly phase change (<j>). Positive phase changes (warm colours) indicate a relative increase in distance to the satellite; (c) Phase profile perpendicular to the North Anatolian Fault (along the dashed line in (b)). The grey bands delimit the 1- and 2-sigma error bounds, with red bars the GPS velocities. Phase changes predicted by an elastic model are plotted as a dashed line.

Fig. 6 (a) Topographic and tectonic map of the eastern end of the North Anatolian Fault. The coloured area is an elevation model calculated from a 1-day interferogram: the North Anatolian Fault (dashed red line) can clearly be seen cutting through the landscape. Arrows are GPS-determined velocities relative to the Eurasian plate (McClusky et al., 2000); (b) Stacked inter-seismic interferogram, converted to a yearly phase change (<j>). Positive phase changes (warm colours) indicate a relative increase in distance to the satellite; (c) Phase profile perpendicular to the North Anatolian Fault (along the dashed line in (b)). The grey bands delimit the 1- and 2-sigma error bounds, with red bars the GPS velocities. Phase changes predicted by an elastic model are plotted as a dashed line.

planet, which could used to construct accurate medium-range earthquake forecasts.

Was this article helpful?

0 0

Post a comment