Search

To better understand the shallow structure of El Hierro island, we determined a 3D S-wave velocity model using Ambient Noise Tomography (ANT). We exploited a dataset recorded by 21 broadband seismic stations deployed at El Hierro island in two surveys realized in 2015 and 2021. This dataset allowed us to obtain empirical Green’s functions by cross-correlating seismic ambient noise signals and retrieving 105 dispersion curves using the frequency-time analysis (FTAN). Then we obtained 2-D Rayleigh wave group velocity maps for periods between 0.6 s and 2.6 s through a non-linear multiscale inversion (Cabrera-Pérez et al, 2021). Finally, we performed depth inversion through a Bayesian transdimensional approach to obtain a 3-D S-wave velocity model. The obtained ANT model is merged with a local earthquake tomography model (García-Yeguas et al, 2014). Our study highlights six relevant seismic velocity anomalies. We observed the presence of three high-velocity zones located in the eastern, western and northern parts of the island, which could be related to intrusive bodies possibly associated with the formation of El Hierro island. We also observed three low-velocity anomalies in the northern and southern parts. The anomaly in the North of the island could be related to loose deposits generated by the El Golfo valley megalandslide. The anomalies in the South could be related to porous and highly fractured materials produced during the more recent volcanic episodes., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

Volcanic eruptions are key scenarios for developing new tools for volcanic monitoring. Among the geophysical techniques, electromagnetic methods are not extensively used in volcano monitoring. However, these methods are very sensitive to temporal changes in the subsoil due to the movement of fluids, which strongly affects the electrical resistivity of the subsoil.During the volcanic eruption that started on the island of La Palma on September 19th, 2021, and the subsequent post-eruptive process, we deployed several magnetotelluric (MT) stations in the vicinity of the new volcano to test the performances of such instrumentation as an effective volcanic monitoring tool. Magnetotelluric stations have been installed for continuous monitoring, recording electric and magnetic fields along the N-S and E-W directions. We have obtained good quality transfer functions for the period range of 0.01 - 100 s. The main objective of this MT experiment was to detect possible variations of the apparent resistivity and phase curves and to compare them with other geophysical and geochemical studies conducted in the same area. The preliminary results show slight changes in the resistivity over time that could be related to the development of a shallow hydrothermal system around the volcano. Furthermore, in comparing the pre-eruptive 3D resistivity model of the island, we observed that regions with low resistivity were almost aseismic, possibly due to the presence of ductile rocks like clays.

The 2021 Tajogaite eruption was characterized by short precursors, lasting only eight days. The seismicity started on Sep. 11th with a westward and upward migration of hypocenters. Permanent GNSS stations started recording deformation of the island's western side, which reached more than 15 cm just before the eruption.After the eruption's onset, which occurred on September 19th, the deformation increased, reaching a maximum on September 22nd and later showing a nearly steady deflation trend in the following months. To better understand the dynamics of the eruption, we exploited a joint dataset of GNSS and Sentinel-SBAS time series along both ascending and descending orbits. To obtain the geometry of the causative source of the ground deformation, we combined the result of a preliminary non-linear inversion and the precise distribution of the seismicity. The resulting geometry is that of a twisted dike bending eastward.We realized a spatio-temporal inverse modelling to obtain the evolution of the causative source of the ground deformation. The inverse modelling has been realized using a 3D finite-element approach considering the island's topography.The main results we obtained are that we observe a correspondence between the magmatic intrusion and pre-eruptive seismicity; the ascent of the magma occurred along two different branches; the rheology of a previously identified ductile layer strongly affected the magma propagation process. Finally, we evidence an early shallow deformation which we interpret as the effect of ascending hydrothermal fluids.Our findings highlight the need for advanced modelling to understand pre-eruptive processes in basaltic volcanoes., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The receiver function analysis (RF) is a well-established method to investigate the crustal and upper mantle structure in different geodynamic settings. It is highly sensitive to seismic discontinuities and provides information about P- and S-wave velocities beneath a seismic station.This work aims to study the crust and the upper mantle of La Palma Island up to 50 km of depth by using RF analysis. La Palma is one of the youngest oceanic volcanic islands of the Canary archipelago, with a complex internal structure that makes applying conventional RF challenging. The island was affected in 2021 by the devastating Tajogaite eruption which occurred on the Cumbre Vieja volcano.For this study, we obtained RFs of five stations using multi-taper deconvolution. After that, we applied the transdimensional approach of Bodin et al. (2012) to determine 1D profiles of P- and S-wave velocities and the probability of a discontinuity beneath each station.Each 1D model was interpolated to build a N-S trending seismic section that allowed us to correlate the different discontinuities. We observe a Low-Velocity Zone in the upper mantle beneath the Cumbre Vieja volcano and four different layers with a significant change in their Vp/Vs ratios. These results are compared with the seismicity recorded beneath the island from 2017 until the pre-eruptive phase of the Tajogaite eruption. Hypocenters are mainly located at the base of the crust (10-15 km deep) and in the upper mantle (20-30 km deep), possibly related to the presence of two subcrustal magmatic reservoirs., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

Time-lapse monitoring of seismic velocity at a volcanic area can provide important information about the dynamics of the volcanic system and its temporal variability. One standard technique to monitor small changes in the medium is ambient noise interferometry. This technique is based on quantifying relative velocity variations (dv/v) by measuring changes in the ambient noise cross-correlation signal phase. In this study, we demonstrate that together with velocity variations, the study of changes in seismic attenuation (Qc) extracted from ambient noise cross-correlation could be relevant for volcano monitoring. The Qcis determined using the lapse-time dependence method (Calvet and Margerin, 2013), in which, Qcis measured as a function of the coda window length for different onsets of the ambient-noise cross-correlation coda. We apply this technique to the 2021 Cumbre Vieja eruption in La Palma (Canary Islands), which started on Sept. 19th and had a significant social and scientific impact. First, we determine a reference Qcmodel of the Cumbre Vieja volcano using data from Aug. 1st to 31st, 2021. Then, we analyze the spatio-temporal Qcvariations during the nineteen days preceding the eruption, and we compare our results with previously obtained dv/v results. We observe an increase of Qcduring the pre-eruptive phase, corresponding to a dv/v decrease. We hypothesize that the observed correlated Qcincrease and dv/v drop could be explained by the ascent of hydrothermal fluids towards the surface before the eruption., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The first continuous automatic geochemical station in Canary Islands was installed at summit cone of Teide volcano (Tenerife island) in 1999. The time series show anomalous changes of the diffuse CO2 emission with values ranging between 0 and 62.8 kgm-2d-1, with a mean value of 4.3 kgm-2d-1.We analyzed the CO2 efflux time series during 1999-2011 by using Fast Fourier Transform (FFT), the Continuous Wavelet Transform (CWT) and Hilbert-Huang Transform (HHT) methods to detect relevant time-frequency patterns in the signal. The spectrum of FFT shows large values at differents periods: 72 days, 6 month, 1, 1.6 and 5-6 years. Aplaying Ricker wavelet, the CWT scalogram shows one of the most striking features is the quasi-periodic oscillation visible between mid-2002 and 2011 at scales around 500 days, related to the increments of filtered CO2 efflux. The oscillation with a period of about 3 years probably reflects an internal and deep modulation of the dynamics of the magmatic-hydrothermal system of Teide volcano. The analysis evidenced also oscillations with a period of about 6 months during the interval 1999-2011. Another relevant component visible in the scalogram is the quaterly oscillation (scales of about 90 days) visible along most of the time series due to not completely filtered climatic modulation. The HHT also reveal a high amplitude at period of 90 days, 6 months, 1 and 5.5 years. The HHT application offers some promising results since it provides more precise estimates due to the adaptability in the time and frequency resolutions, even with noisy signals., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

In recent years the use of Distributed Acoustic Sensing (DAS) in seismology is gaining extensive usage in different applications. A High-Fidelity DAS system (HDAS) was deployed during the 2021 Tajogaite eruption on Cumbre Volcano (La Palma, Canary Islands), allowing the recording of most of the syn-eruptive and post-eruptive seismicity. The eruption lasted from Sep. 19th until Dec. 13th of 2021. The HDAS was installed on 19th Oct. 2021 and is still operating.The HDAS was installed around 10 km from the eruptive vent and was connected to a submarine fibre optic cable directed toward Tenerife Island. Since then, the HDAS has been recording seismic with a temporal sampling rate of 100 Hz and a spatial sampling rate of 10m for a total length of 30 (first phase) and 50 km using Raman Amplification (last period).The HDAS recorded thousands of local earthquakes as well as regional and teleseismic events. It was revealed to be an excellent tool for volcanic monitoring, allowing a better location of deeper events which location was made difficult by the small aperture of the seismic network of La Palma.The HDAS was also able to record the low-frequency (, The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

In the last decades, the magnetotelluric (MT) method has been proved to be a useful geophysical tool in different contexts, from geothermal reservoir characterization to crustal structures studies. Nevertheless, the MT method is very sensitive to the presence of noise. Indeed, as the method is based on the measurement of both the electric and magnetic components of the natural electromagnetic (EM) field, it fails when these components are affected by noises of different nature. In particular, in industrialized and urbanized context, the MT time series could be strongly affected by man-made noise and, as a consequence, the impedance tensor estimates, given by the ratio between the electric and magnetic components of the MT field, could be unreliable. To improve the reliability of these estimates, most of the proposed approaches rely on the robust evaluation of the impedance tensor as well as on the use of remote reference MT stations or on the combination of both approaches. However, these methods are not always effective. The robust methods fail when most of the data are affected by noise, giving as result a biased impedance tensor, while the remote reference approach is ineffective when the noise is correlated between reference and local MT station. In recent years, alternative procedures have been proposed to obtain reliable estimates of the MT impedance tensor. In the present work, a different approach based on the use of Discrete Wavelet Transform (DWT) and Self-Organizing Map (SOM) neural network analysis is proposed for improving the magnetotelluric impedance tensor estimates. The approach has been tested by changing type, level and window length of the noise affecting MT time series. Furthermore, in order to identify the most reliable apparent resistivity and phase values among the different impedance tensors clusters provided by the SOM analysis for each analyzed period, a selection criterion is provided and tested on synthetic and field MT data.

Magnetotellurics (MT) is a geophysical method based on the use of natural electromagnetic signals to define subsurface electrical resistivity structure through electromagnetic induction. MT waves are generated in the Earth’s atmosphere and magnetosphere by a range of physical processes, such as magnetic storms, micropulsations, lightning activity. Since the underground MT wave propagation is of diffusive type, the longer is the wavelength (i.e. the lower the wave frequency) the deeper will be the propagation depth. Considering the frequency band commonly used in MT prospecting (10-4 Hz to 104 Hz), the investigation depth ranges from few hundred meters to hundreds of kilometers. This means that magnetotellurics is inherently a multiscale method and, thus, appropriate for applications at different scale ranging from aquifer system characterization to petroleum and geothermal research. In this perspective, the application of the Wavelet transform to the MT data analysis could represent an excellent tool to emphasize characteristics of the MT signal at different scales. In this note, the potentiality of such an approach is studied. In particular, we show that the use of a Discrete Wavelet (DW) decomposition of measured MT time-series data allows to retrieve robust information about the subsoil resistivity over a wide range of spatial (depth) scales, spanning up to 5 orders of magnitude. Furthermore, the application of DWs to MT data analysis has proven to be a flexible tool for advanced data processing (e.g. non-linear filtering, denoising and clustering).