Interpreting seismicity and earthquake locations during the recent unrest on the Reykjanes Peninsula in SW-Iceland

The Reykjanes Peninsula in SW Iceland is transacted by a divergent plate boundary with oblique spreading. Volcanic unrest periods, marked by fissure eruptions widely across the peninsula, seem to occur with regular intervals, approximately every 800-1000 years. These volcanic unrest periods have durations of 100 up to 400 years. In 2020, it became clear that magma was on the move again after 800 years of quiescence, as repeated uplift was measured in the vicinity of the Svartsengi geothermal area. Minor subsidence was recorded between uplift periods and seismicity increased again, only after the previous state of uplift had been surpassed, in line with the so-called Kaiser effect. A year later, a dike intrusion in Fagradalsfjall triggered earthquake activity tens of km away as stored tectonic stresses along the peninsula were released. After a clear decline in earthquake activity, an eruption took place in Geldingdalir, Fagradalsfjall, on 19 March 2021, the first one in over 6000 years in that region. At the time of writing, 4 volcanic eruptions have occurred since 2021 and in total roughly 20 magmatic intrusive events have taken place on the Reykjanes Peninsula. Recently an escalation in volcanic activity has been observed. In late October 2023, the 5th period of uplift started in Svartsengi signifying faster magma inflow rates than previously inferred. Seismicity increased and was widespread in line with increased stresses above an inflating sill at about 5 km depth. On 10-11 November, during nearly 12 hours of intense seismic activity, the magma found its way from the magma storage beneath Svartsengi some 2 km laterally towards the center of an old crater row and creating a 15 km long shallow dike. Subsidence was observed above Svartsengi as the magma was drained from beneath and a graben formed beneath the coastal town of Grindavk where extensive faulting caused considerable damage. On 18 December, a similar but smaller magma intrusive event originating in Svartsengi occurred, causing an eruption approximately at the center of the original dike. This time, earthquakes only occurred about 90 minutes before the eruption onset and no clear trend of earthquakes migrating from Svartsengi towards the laterally offset dike were detected. At the time of writing (10 January, 2024), a similar amount of magma volume is inferred to have accumulated beneath Svartsengi since shortly before the last eruption, however, seismicity is still at normal background levels. The volcano monitoring team at the Icelandic Meteorological Office in close collaboration with geoscientists at the Insitute of Earth Science at the University of Iceland and HS Orka, have been under immense pressure to interpret the ongoing activity. A vital part has been to interpret seismicity rates and earthquake locations and any changes thereof, along with modeling dike and sill inflow rates from geodetic measurements. We show that meaningful interpretation of earthquake activity can only be done when jointly interpreted together with deformation and stress models as stress changes heavily influence earthquake locations and the temporal onset of earthquake activity.