In the South Iceland Seismic Zone (SISZ), the recent and present-day tectonic mechanisms are studied by geological and geophysical means. The determination of the stress regimes is done through calculation of stress tensors, involving the use of inverse methods. Data inversion is applied in a nearly identical way to fault slip data and to sets of double-couple focal mechanisms of earthquakes.

As pointed out before (Angelier et al., 1996, in " Seismology in Europe ", European Seismological Commission, Reykjavik, p.199-204), preliminary analyses of the focal mechanisms recorded by the S.I.L. network of the Icelandic Meteorological Office revealed the presence of two contrasting stress regimes.

A major NW-SE extension is in agreement with the general left-lateral behaviour of the E-W trending SISZ. A minor but significant NE-SW extension is attributed to rebound phenomena and local block accommodation. For both these regimes, strike-slip mechanisms prevail, but many normal and few reverse mechanisms are also present. Most normal mechanisms fit with the NW-SE extension, whereas the reverse ones are rather compatible with the secondary tectonic regime. Two main types of permutations between stress axes, / and / , are common.

However, a crucial problem, which does not exist while studying faults in the field, is raised in the analysis of double-couple focal mechanisms of earthquakes. It deals with the choice between nodal planes. This problem is solved in two ways.

First, some methods do not require the choice between the nodal planes: this is the case for the right dihedra, or P- and T- dihedra, method (see Vordufell examples in Angelier et al., 1996) and for two new geometrical and numerical methods (in prep.).

Second, the same inversion methods used for fault slips are applicable to double-couple earthquake mechanisms, provided that the choice between the nodal planes is done. Because in the SISZ one deals with very shallow earthquakes, we claim that even in a seismological study, one must take advantage of the rich geological information available. We consequently propose to perform these choices between nodal planes according to an integrated approach including four criteria, geophysical and geological: (1) the consideration of the nodal plane attitude (strike and dip), as compared with that of all geological faults and planes of mechanical weakness known from field observation at sites, geological mapping and aerial photograph analyses; (2) the comparison with the recent fault mechanisms observed in the field (including pitch and sense of motion as well as strike and dip of fault); (3) the mechanical likelihood of each of the two fault solutions, based on taking friction into account (e.g., a shallow dip is more likely than a steep one for a reverse fault); (4) the best fit criterion relative to the stress tensors calculated. If external (i.e., comparisons with geological tensors), it is of good value. If internal (i.e., the nodal planes resulting in the best possible fit within the data inversion process), it is somewhat circular, hence disputable.

An other important aspect in such inversion of earthquake mechanisms is the selection and weighting process taking into account the multiple characteristics of double-couple solutions (reliability, magnitude, depth, etc.)

The systematic use of the four criteria for nodal plane selection within a weighted approach allowed more accurate determination of the stress regimes in the SISZ and better understanding of the geological significance of the earthquake mechanisms, although the general trends mentioned before remain valid.