Nature research paper: Deformation and seismicity decline before the 2021 Fagradalsfjall eruption
, allowing for opening and shear, and the observed movement at the central axis of the plate boundary with a second dislocation, also allowing for opening and shear. We modelled the surface displacements resulting from the two largest tectonic earthquakes as separate rectangular dislocations, allowing for shear motion only.
The output parameters from this inversion confirmed that a dyke was intruding along vertical planes defined by the seismicity. In a second inversion, we substituted the single dyke dislocation for two connected dislocations with strikes fixed to that indicated by the relocated seismicity .
For our final inversion we fixed the location of the dyke to the preferred location from the second inversion with maximum lengths and depths derived from the seismicity. We also fixed the location, dip and strike of the plate boundary dislocation from the second inversion. Very little opening was assigned to this segment in the second inversion, so we allowed for shear motion only.
The trade-off between daily volume change and mean depth of magma emplacement was studied with a simulation test. An incremental dyke opening inversion was carried out on the simulated data to test whether or not the observed trade-off between depth and volume change for the daily solutions influences the overall relationship between depth and volume change.
. Despite the trade-off between depth and volume for individual days, the simulated constant volume change rate is retrieved, with no dependence on depth.≥ 4, except for the two largest events on 24 February and 14 March, was evaluated by the following procedure. Each earthquake was assumed to be on a rectangular north–south striking vertical strike-slip fault in accordance with the dominating fault mechanisms.