Task 3: Comparison between global earth models, including sphericity and self-gravitation, and plane models

Start: March 1996 (month 1)
End: October 1997 (month 20)
Responsible partner: UBLG.DF

Major results on this topic have been presented in a paper entitled: Post-seismic deformations: a comparison between spherical and flat earth models and stress diffusion following large earthquakes, in press on Geophysical Journal International []. In this work we have compared the co- and postseismic deformations of a spherical and of a flat model perturbed by strike-slip earthquakes. A new manuscript is being submitted, Spherical vs. flat models of postseismic deformation [], where a more detailed study is performed for general shear dislocations. The purpose of these investigations is to provide a clear account of the effects of sphericity on postseismic motions, and to define the range of validity of flat models currently employed in the literature. An excerpt of the results obtained are shown in Figure 32.

**Figure:** *See text for discussion.*
$\begin{figure}% latex2html id marker 2580 \refstepcounter{myfigure} \begin{cente... .../bonafede/figure-3.eps, angle=-90, width=0.9\textwidth }\end{center}\end{figure}$

In this case, we show the displacement in the near field (d=100 km) and in the far field (d=1000 km) induced by a vertical strike-slip fault with width of 50 km buried in the lithosphere at a depth of 25 km. The displacement u along the strike of the fault is evaluated as a function of time for two models: a spherical and a flat one. The former has been first proposed by Piersanti et al. [], whereas the latter has been introduced by Bonafede et al. []. In both cases we account for the delayed response of the asthenosphere. At any time t (given in units of Maxwell times of the asthenosphere), the displacement predicted by the spherical model is sensibly larger than the one provided by the flat one. The discrepancies increase with increasing time, up to a factor of 2 in the case of far-field deformations. These results reveal that a precise determination of postseismic motions must be based on spherical models even when relatively short source-observer distances are involved. We have verified that this also holds for the residual time-dependent stress field following both strike-slip and dip-slip earthquakes. This could have implications on the interpretation of the process of stress diffusion and particularly on the interaction between postseismic stress field and the evolution of the Mid-Atlantic Ridge in the Iceland region.