No abstract

Taupō volcano, New Zealand, is a large caldera volcano that has been highly active through the Holocene. It most recently erupted ∼1,800 years ago but there have been multiple periods of historic volcanic unrest. We use seismological and geodetic analysis to show that in 2019 Taupō underwent a period of unrest characterized by increased seismic activity through multiple swarms and was accompanied by ground deformation within the caldera. The earthquakes, which include non‐double‐couple events, serve to outline an aseismic zone beneath the most recent eruptive vents. This aseismic zone is coincident with an inflating source, based on forward modeling of ground deformation data. We infer that this aseismic and deforming region delineates the location of the present day magma reservoir that is ≥250 km3 in volume and has a melt fraction of >20%–30%, inhibiting seismic activity. Our analysis shows that the 2019 unrest at Taupō was volcanic in nature and origin, demonstrating that this is an active and potentially hazardous volcano, and that improving our monitoring and understanding of its behavior is important.

Silicic caldera volcanoes are frequently situated in regions of tectonic extension, such as continental rifts, and are subject to periods of unrest and/or eruption that can be triggered by the interplay between magmatic and tectonic processes. Modern (instrumental) observations of deformation patterns associated with magmatic and tectonic unrest in the lead up to eruptive events at silicic calderas are sparse. Therefore, our understanding of the magmatic-tectonic processes associated with volcanic unrest at silicic calderas is largely dependent on historical and geological observations. Here we utilize existing instrumental, historical and geological data to provide an overview of the magmatic-tectonic deformation patterns operating over annual to 104 year timescales at Taupō volcano, now largely submerged beneath Lake Taupō, in the rifted-arc of the Taupō Volcanic Zone. Short-term deformation patterns observed from seismicity, lake level recordings and historical records are characterized by decadal-scale uplift and subsidence with accompanying seismic swarms, ground shaking and surface ruptures, many of which may reflect magma injections into and around the magma reservoir. The decadal-scale frequency at which intense seismic events occur shows that ground shaking, rather than volcanic eruptions, is the primary short-term local hazard in the Taupō District. Deformation trends near and in the caldera on 101–104 yr timescales are atypical of the longer-term behavior of a continental rift, with magma influx within the crust suppressing axial subsidence of the rift basin within ∼10 km of the caldera margin. Examination of exposed faults and fissures reveals that silicic volcanic eruptions from Taupō volcano are characterized by intense syn-eruptive deformation that can occasionally extend up to 50 km outside the caldera structure, including ground shaking, fissuring and triggered fault movements. We conclude that eruption and unrest scenarios at Taupō volcano depend on the three-way coupling between the mafic-silicic-tectonic systems, with eruption and/or unrest events leading to six possible outcomes initially triggered by mafic injection either into or outside the magma mush system, or by changes to the tectonic stress state.

Tarawera volcano (New Zealand) is volumetrically dominated by rhyolitic lavas and pyroclastic deposits, but the most recent event in AD 1886 was a basaltic Plinian fissure eruption. In March 2019 a swarm of at least 64 earthquakes occurred to the NE of Tarawera volcano, as recorded by the New Zealand Geohazard Monitoring Network (GeoNet). We use seismological analysis to show that this swarm was most likely caused by a dyke that intruded into the brittle crust between depths of 8–10 km and propagated toward Tarawera volcano for 2 km at a rate of 0.3–0.6 m s−1. We infer that this was a dyke of basaltic composition that was stress-guided toward Tarawera volcano by the topographic load of the volcanic edifice. Dyke intrusions of this nature are most likely a common occurrence but a similar process may have occurred during the 1886 eruption with a dyke sourced from some lateral distance away from the volcano. The 2019 intrusion was not detected by InSAR geodesy and we use synthetic models to show that geodetic monitoring could only detect a ≥6 m wide dyke at these depths. Improvements to geodetic monitoring, combined with detailed seismological analysis, could better detect future magmatic intrusions in the region and serve to help assess ongoing changes in the magmatic system and the associated possibilities of a volcanic event.

No abstract