Ragnar Stefánsson, Kristján Ágústsson and Páll Halldórsson.
In the paper "Strain release and strain build-up in the south Iceland seismic zone" by Stefánsson, R. and Halldórsson, P., Tectonophysics 1988, they discussed and constrained two models to explain different features observed for the South Iceland Seismic Zone, SISZ.
The first model is that of strain build up by shearing and release by earthquakes of the zone as a whole. It was constrained by: On one hand 2 cm /year drift divergency taken upp by transform motion along the SISZ and shearing it as a whole. This is compared with moments calculated from surface wave magnitudes, Ms, estimated by the empirical formula of Purcaru and Berckhemer, Tectonophysics, 1978: log Mo=16.1+1.5Ms. Rigidity was based on body wave velocities and the thickness of the cust involved in the build up of stress was assumed 15 km. The agreement between moment build up and release was good for the period since 1700, the period of well described earthquakes. In fact the agreement of this model was too good. It was noted by the authors, that in fact the earthquake energy was too large compared to the energy build up of normal drift velocity as a significant part of the divergency moment should be taken up aseismically. Also they noted that, especially in the western part, the involved elastic-brittle crust is much thinner than the 15 km thickness assumed in the model.
As a suggestion to explain that we have more earthquake energy in the seismic zone than that expected from 2 cm yearly drift, it is pointed out that it is physically plausible that the build up of strain energy in the system is not only due to shearing because of rift divergency but energy from magmatic fluid intrusions should be added to this as well as in general the energy of variable push from the hot spot.
The other model has to do with the characteristics of individual earthquake faults in the SISZ. This is the "dual model of earthquake mechanism" suggested in the paper to explain the right lateral motion on NS trending fault planes of the individual historical earthquakes in this zone of general EW left lateral transform motion. The model was suggested in light of the interplay between the NE-SW compression exerted by the plume and the general almost EW plate divergency in this region. The model is that of strike slip earthquakes governed by compression from the Iceland hot spot. The earthquake slips were proposed to release the compressional stresses so rotation of the principal stresses occurs, so the maximum compression comes closer to NS, and the least compression closer to EW. Thus the side pressures on the NS faults would be lowered which temporarily opens for fluid intrusions into NS cracks. This also explains why the observed NS earthquake faults on the surface are not repeated during the 300 years of earthquake history, but rather faults 5 - 10 km away. The intrusions increase the compression in their surrounding, closing the nearby cracks for a long time.
The recent modelling of the Vatnafjöll earthquake sequence of 1987 is that of foreshocks and a 5.8 earthquake followed by magma intrusion into pure strike slip fault break. We thus have two active processes with a different time scale, one of discrete slips measured by seismometers, the other is that of slow expansion measured by volumetric strainmeters. The epicenter of the Vatnafjöll earthquake is on the eastward prolongation of the SISZ, within 10 km of what usually is considered to be its eastern end. The fault plane of the earthquake was near vertical in NS direction and the slip right lateral (Ágústsson, K. et. al 1997).
To say that such a model is aplicable also to the SISZ is not certain. Since several eruptions have occurred in the Vatnafjöll area, it is probable that it has some kind of magma chamber or a magma layer. But it has also been claimed on various reasons that there are fluids below the whole of the SISZ. As pointed out earlier in this abstract, excess earthquake energy in SISZ may be explained by energy from magmatic fluid intrusions. It is at least possible that the Vatnafjöll earthquake model, i.e. that of fault break followed by fluid intrusion, comparable with the dual mechanism model, is valid for the whole of the SISZ.
In general it has been claimed, especially on basis of magnetotelluric measurements that fluid mobility, especially that of water fluids, exist bellow the crust or in the lower part of the crust all over the world. It is thus possible that the Vatnafjöll model is generally relevant although the process of fluid intrusions plays a more significant role in volcanic than in nonvolcanic areas.