Magmatic connectivity among six Galápagos volcanoes revealed by satellite geodesy

Shallow magmatic reservoirs that produce measurable volcanic surface deformation are often considered as discrete independent systems. However, petrological analyses of erupted products suggest that these may be the shallowest expression of extensive, heterogeneous magmatic systems that we show may be interconnected. We analyse time series of satellite-radar-measured displacements at Western Galápagos volcanoes from 2017 to 2022 and revisit historical displacements. We demonstrate that these volcanoes consistently experience correlated displacements during periods of heightened magma supply to the shallow crust. We rule out changes in static stress, shallow hydraulic connections, and data processing and analysis artefacts. We propose that episodic surges of magma into interconnected magmatic systems affect neighbouring volcanoes, simultaneously causing correlations in volcanic uplift and subsidence. While expected to occur globally, such processes are uniquely observable at the dense cluster of Western Galápagos volcanoes, thanks to the high rate of surface displacements and the wealth of geodetic measurements.

Fig. S1 Descending Sentinel-1 data for the Western Galápagos.a, Time series of displacement of the reference pixel, showing its variation with time.b-g, Time series of displacement at each of the major volcanoes of the Western Galápagos, from 07/01/2017-30/03/2022.Annotated are known periods of significant unrest, including unrest at Cerro Azul, and eruptions at each of Fernandina, Sierra Negra, and Wolf.The grey areas denote the periods used for Independent Component Analysis, as presented in this study.h, Wrapped cumulative displacement map of the Western Galápagos, across the entire time series.Each fringe corresponds to 10 cm of range change in the satellite line-of-sight.The arrow shows the satellite heading, as well as the average incidence angle.The annotated points refer to pixels used during correlation analysis, while the reference area is used during Independent Component Analysis (Figure 2).Fig. S2 Results of cross-correlation analysis for the entire time series of each volcanic pair, in Descending.Similarity between time series is presented on the y-axis, while time lag is presented on the X.The time lag refers to satellite acquisitions between time series points being compared (e.g. a lag of five means there was five satellite acquisitions between points).The correlation coe cient for each pair of time series is annotated on the right y-axis.

Fig. S3
Fig. S3 Time series of displacement at each volcano in the Western Galápagos from 1998-2011, modified from [S17].In each case, dots represent vertical displacement, in centimetres, while the vertical lines represent various eruptions.The erupted volcano is denoted by colour, while the episodes of correlated displacements observed by [S17] are marked by vertical green lines.These data are a compilation from multiple satellite missions (ERS-1/2, Envisat, Radarsat, AlOS-1.)

Fig. S6
Fig.S6Influence of choice of reference pixel, at each volcano.In each case, the plotted pixel is held constant.Descending and Ascending time series are presented in each column, while a di↵erent volcano is plotted in each row.Line colours correspond to the plotted reference pixel, the location of which is presented in the inset map.

Fig. S7
Fig. S7 Change points at each Western Galápagos volcano in the ascending track direction, from 2017-2022, as identified on the basis of change in rate or direction.Change points are annotated in red, while the grey boxes illustrate a one month period either side of the change point.

Fig. S8
Fig. S8 Change points at each Western Galápagos volcano in the descending track direction, from 2017-2022, as identified on the basis of change in rate or direction.Change points are annotated in red, while the grey boxes illustrate a one month period either side of the change point.

Fig. S9
Fig. S9 Dilational strain associated with the 2018 eruption of Sierra Negra, modelled using Coulomb 3.1 [S48].The eruption is approximated as a point source with a volume decrease of 2x10 8 m 3 [S30]

Fig. S10
Fig.S10Periods of heightened correlation between Western Galápagos volcanoes, determined using windowed correlation analysis, using a window size of 150 days.Red lines illustrate a significant unrest event (e.g.eruption).a-b, Pairwise correlation coe cients for Western Galápagos volcanoes.The location of the pixel used to create each time series is illustrated in Figure1.Vertical lines and dots at each volcanic pair denote a respective correlation coe cients of >| 0.9 | and >| 0.75 | across the corresponding window, with each point plotted at the centre of the window.Horizontal lines illustrate the temporal extent of each window.c, Stacked area plot of correlation coe cients of >| 0.9 | to illustrate periods of heightened correlation.These identified periods are separated by grey boxes.

Fig. S12
Fig. S12 E↵ect of GACOS correction on each data set.The left columns plot the mean standard deviation before correction against that after correction, while the right columns plot the percentage change in standard deviation due to the correction.

Fig. S13
Fig.S13Periods of heightened correlation between Western Galápagos volcanoes, determined using windowed correlation analysis.Here, with a window size of 15 acquisitions (approximately 3 months, the extent of which is shown by the horizontal black line), to show the a↵ect of alternating the window size.

Fig. S14
Fig.S14The downsampled descending dataset used for Independent Component Analysis.Each point corresponds to the location of one time series of displacement

Fig. S15
Fig. S15 Data, model, and residual for the best fitting sill source at Alcedo, from 26/12/2016-01/10/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.

Fig. S16
Fig. S16 Data, model, and residual for the best fitting sill source at Cerro Azul, from 02/02/2020-03/01/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.

Fig. S17
Fig. S17 Data, model, and residual for the best fitting sill source at Darwin, from 01/07/2020-28/03/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.

Fig. S18
Fig. S18 Data, model, and residual for the best fitting sill source at Fernandina, from 01/07/2020-28/03/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.

Fig. S19
Fig. S19 Data, model, and residual for the best fitting sill source at Sierra Negra, from 29/12/2020-03/02/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.

Fig. S20
Fig. S20 Data, model, and residual for the best fitting sill source at Wolf, from 01/06/2020-27/05/2021.The optimal source parameters, as well as the 2.5% and 97.5% bounds are presented in the underlying table.