Conference Agenda

The Western Galápagos are home to six major volcanic centres that routinely experience geodetically observable ground deformation. Here, deformation varies from high-magnitude uplift at Sierra Negra, long-term subsidence at Alcedo, to frequent co-eruptive unrest at Fernandina. This diverse behaviour, over a compact area (there is approximately 100 km separating the northernmost and southernmost volcanoes), as well as high phase coherence, and limited ground-based monitoring, make the Galápagos well suited to monitoring by space-borne interferometric platforms. We consider ERS and ENVISAT data 1992–2010, and Sentinel-1 data from 2015–present. Here, we construct time series of deformation at each major Galápagos volcano, using automatically produced Sentinel-1 interferograms (using LiCSAR). There were 5 eruptions, at 3 volcanoes (Fernandina, Sierra Negra, and Wolf) during the studied period, as well as a failed eruption at Cerro Azul. We measure continuous volcanic displacements during periods of extrusive quiescence, and observe novel deformation behaviour at each volcano.

Using these Sentinel-1 time series, we observe correlated changes in the magnitude and rate of ground displacement, between neighbouring volcanoes (e.g. multiple volcanoes begin to uplift, simultaneously). We validate these observations by performing Correlation Analysis, as well as Independent Component Analysis, on the constructed time series. We use geodetic source models to estimate the total magma volume flux over the studied period, and show that periods of increased volume flux correspond to periods of correlated deformation, during unrest, resurgence, and eruption. This analysis, suggests that volcanic deformation in the Galápagos is controlled by a common, bottom-up, magmatic origin.

Though broad magmatic processes cause correlated deformation, and control general displacement trends, unrest in the shallow crust and on the surface promotes distinct behaviour at each volcano. Magma degassing at Alcedo drives subsidence in the hydrothermal zone in the southwestern portion of the Caldera. Magma intrusion at Darwin caused uplift in 2020, at an otherwise inactive volcano. The southeastern flank of Cerro Azul uplifted in 2017, due to lateral magma intrusion along an established conduit. There is evidence of an offshore eruption accompanying the 2020 eruption of Fernandina. At both Sierra Negra and Wolf volcanoes, we observe subsidence of lava flows that were emplaced in 2018, and 2015, respectively. At Sierra Negra, we also observe subsidence due to the cooling and crystallisation of a co-eruptive sill, emplaced on the northwestern flank of the volcano.

These observations and analysis of InSAR data unveil complex volcanic deformation in the Galápagos. Bottom-up magma flux controls the major trends in deformation, and can affect multiple volcanoes simultaneously, though shallower processes driven by magmatic fluids can be discerned at each individual volcano.

Unrest episodes at volcanic systems are frequently associated with ground displacements produced by the pressure changes inside magma chambers at depth. The interactions with secondary deformation sources like other magma bodies, shallow geothermal systems, or preexisting tectonic structures like faults can complicate the deformation pattern at the surface. Calderas are volcanic systems characterized by a large depression formed after a magma chamber roof collapse and bordered by ring faults. We can thus expect stress interactions between the magma source and the ring faults at calderas during unrest episodes or eruptions. This indeed has been occurring at Askja volcano since the current unrest episode started in August 2021.

The Askja volcanic system is located in the North Volcanic Zone of Iceland. It consists of a central volcano with three nested calderas, Kollur, Askja, and Öskjuvatn, and fissure swarms to the north and southwest. Geodetic data (leveling, GNSS, and InSAR) have shown that the Askja caldera floor continuously subsided since 1983. Then, in August 2021, the volcano entered a period of unrest with rapid uplift (~3 cm/week) and increased seismic activity below the lake that fills the smallest and youngest Öskjuvatn caldera (formed after the 1875 eruption). As of early March 2023, GNSS data show ongoing uplift, with the daily earthquake count higher than before the summer of 2021. Furthermore, satellite optical images from early 2023 show that the ice cover on Öskjuvatn lake, which generally lasts until early summer, has completely melted, indicating increased water temperature in the lake.

Here we use Sentinel-1 SAR images acquired from four different orbits (two ascending and two descending) to study the ground deformation at Askja volcano between 2016 and 2022. Only images acquired in late Summer and early Fall can be used since the area is covered by snow during the rest of the year, preventing retrieval of the deformation signal due to lack of coherence. Our InSAR time series results show that the Askja caldera floor subsided with a steady rate of ~1.5 cm/yr between July 2016 and July 2021, and then, in early August 2021, the displacement changed to uplift. In only one month, the uplift matched the subsidence of the previous five years. By September 2022, the maximum uplift had reached ~40 cm, near the western shore of the Öskjuvatn lake, close to the center of the larger Askja caldera. Interestingly, the deformation maps for all four orbits show an asymmetric pattern that follows the ring faults in the northwestern part of Askja caldera. The pattern is somewhat similar for both the subsidence and uplift periods. This suggests that the same magma body was deflating, prior to the unrest, and then inflating when the pressure started to increase in August 2021. We use analytical models to evaluate the inflating source parameters. The best model is a sill-like source, NW-SE elongated (5.5 x 2.5 km²), located at ~2.5 km depth below the surface. However, with this one source, we could explain less than 50% of the observed deformation. Therefore, with boundary element modeling, we first introduce the Askja caldera ring faults into our model setup. While the magmatic intrusion accounts for the broad uplift, the ring-fault movement localizes the deformation close to the caldera rim, yielding a better fit to the observed data. Furthermore, by adding the ring faults of Öskjuvatn caldera as well, we introduce an asymmetry that further mimics the observed data. The Askja volcanic system experienced a similar unrest episode in the 1970s. For two years, uplift was observed by leveling and may have continued for a third year. This episode did not lead to an eruption. Whether the current unrest period leads to an eruption or not, we cannot say, and at the moment, InSAR is unusable to monitor the volcano due to snow. However, GNSS and seismic data are used to detect any changes in the volcano behavior that could indicate the end of the unrest episode or an imminent eruption.

ISVOLC is a 10 partner research project funded by the Icelandic Research Fund, addressing the effects of climate change-induced ice retreat on seismic and volcanic activity. The project start date is 1 April 2023, and it has a duration of 3 years. The project is led by the Icelandic Meteorological Office, together with the University of Iceland.

Glaciers in Iceland have been retreating since 1890 and climate change simulations predict that the majority may disappear within a few hundred years. Retreating ice caps change the subsurface stress field. Glacier covered volcanic systems are most affected, but also crustal conditions outside glaciers. Eruption likelihood may be modified, as occurred during the Pleistocene deglaciation. More melt is estimated to form under Iceland because of ice retreat. However, there are several uncertainties: i) if, how and when this new magma reaches the surface; ii) if stability of existing magma bodies is modified; iii) if deglaciation is already resulting in accumulation of larger volumes of melt within crustal reservoirs; iv) how induced variations in the stress field may affect both future volcanic and seismic activity. ISVOLC will address these research questions using four active volcanoes (Katla, Askja, Grímsvötn and Bárðarbunga) and two major fault zones (South Iceland Seismic Zone and Tjörnes Fracture Zone) in Iceland, that serve as a natural laboratory for studying the effects of deglaciation on volcanism and seismicity.

The project will generate new Glacial Isostatic Adjustment (GIA) models including estimates of magma generation and 3D finite element models of magmatic plumbing systems beneath the target volcanoes. Combined crustal stress changes from GIA and magma movements will be used to infer the influence on stability of existing magma bodies beneath these volcanoes and for determining the effect on fault zones. Simulated scenarios of continued ice mass loss will be used to assess future changes in volcanic and seismic activity, for improved understanding of natural hazards.

This presentation will provide an overview of the project and how Earth observation data will be utilised to achieve the project objectives. An update on the status of the target volcanoes will also be presented, including the latest satellite interferometry and geodetic modelling results.

Tulu Moye is an actively deforming volcanic complex with a geothermal field in the central part of the Main Ethiopian Rift (MER). We use ascending (087) and descending (079) tracks of Sentinel-1A/B derived InSAR data between 2014 and 2022, integrated with other geophysical data, to investigate the temporal and spatial characteristics of the deformation signal in the area, and to model its source. Interferograms and time-series analysis show a deformation signal consistent with uplift at a velocity of up to 50 mm/yr in the satellite line-of-sight (LOS) in 2014-2017, then decreasing to 12 mm/yr until 2022. The center of deformation is located about 10 km west of a main geothermal drilling site at Tulu Moye, between the Bora, Berecha, and Tulu Moye volcanoes, with a NW-SE elongation direction. We modelled the source of deformation by jointly inverting the InSAR velocity maps from both tracks through a Monte-Carlo simulated annealing algorithm and then a derivative-based method (quasi-Newton). For the modelling, we assumed a uniform rectangular dislocation sill model, the Okada tensile dislocation, in a conventional elastic half-space. Our best-fit model for the 2014-2017 signal suggests that the deformation is caused by an 8.7 km by 1.2 km sill situated ~7.7 km below the surface (~5.9 km below sea-level), elongated in the N54°W direction and dipping S11°W, and experienced an average rate of volume change ~8.7´10⁶ m³/yr in 2014-2017. The surface projection of the sill overlaps with local transverse faults and hydrothermal manifestations. Furthermore, the sill is ~1-2 km below clusters of microseismic swarms and a region of high resistivity, both indicating hydrothermal fluid flow focused above the sill. The location and geometry of the sill correlates with the upper edge of a high conductivity zone interpreted as partial melt, and we therefore attribute the uplift at the Tulu Moye volcanic complex to inflow of magma in the sill. The inferred inflating sill is below the surface trace of NW-trending faults, caldera rims and surface hydrothermal manifestations, indicating the orientation of the modelled sill may be structurally controlled. We also suggest that these transverse caldera rims and faults may restrict magma flow, and also facilitate both vertical and lateral hydrothermal fluid flow.