Ivan Koulakov

Mutnovsky Volcano, ∼70 km SW of Petropavlovsk‐Kamchatsky, exhibits persistent degassing, hosts active hydrothermal springs and the Mutnovsky Geothermal Power Plant (MGPP) on its slopes. Using the network covariance matrix approach, we analyzed seismic data recorded in 2023–2024 by 15 seismic stations operating around Mutnovsky Volcano. We detected seismic tremor in the 1.5–5 Hz frequency range beginning on 6 February 2024, ∼10 days before a series of hydrothermal explosions occurred in the vicinity of the MGPP. The timing of these explosions was refined based on satellite imagery. Most of the tremor sources were located beneath the northeastern slope of the volcano at shallow depths. The spatiotemporal evolution of tremor activity, alongside independent volcanological data, suggest it was generated by magma‐hydrothermal processes activated by a magmatic fluid influx, which caused the hydrothermal explosions. Thus, observations of such seismic tremors may provide a short‐term precursor for future hydrothermal explosions.

Central and South Kamchatka, separated by the Malko–Petropavlovsky Fault Zone, exhibit a complex geological structure shaped by the long-lived Pacific Plate subduction and associated magmatic and tectonic processes. Magmatism in the region is diverse in both composition and style, expressed through active and dormant volcanic complexes, monogenetic cones, calderas, and geothermal systems. However, the crustal structure governing this activity remains poorly resolved. We compiled three-component seismic records spanning 12 years from a permanent seismic network and six temporary deployments to perform ambient-noise tomography, deriving the shear-wave velocity (Vs) structure of the region. Using both vertical- and horizontal-component noise cross-correlation functions, we measured phase and group interstation velocities of the fundamental modes of Rayleigh and Love waves. To obtain robust phase-velocity estimates from horizontal-component cross-correlations, we expended and validated the Stockwell–Bessel transform method. We then inverted the extracted dispersion curves using a direct three-dimensional Bayesian Monte Carlo approach, obtaining Vs distribution and uncertainties down to a depth of 55 km. Our model images several major active fault zones and reveals previously unidentified fault structures. Beneath active volcanoes, we identify multilayered magmatic systems spatially associated with fault zones. Moreover, we imaged previously unreported magmatic structures beneath the inactive Opala and Vilyuchinsky volcanoes, as well as beneath the Sredinny Ridge. These results provide valuable insights into the tectonic framework, magmatic systems, and their interactions, illuminating the crustal sources of recent volcanism in Central and South Kamchatka.

Mutnovsky Volcano, ~70 km SW of Petropavlovsk-Kamchatsky, exhibits persistent degassing, hosts active hydrothermal springs and the Mutnovsky Geothermal Power Plant (MGPP) on its slopes. Using the network covariance matrix approach, we analyzed seismic data recorded in 2023-2024 by 15 seismic stations operating around Mutnovsky Volcano. We detected seismic tremor in the 1.5-5 Hz frequency range beginning on February 6, 2024, ~10 days before a series of hydrothermal explosions occurred in the vicinity of the MGPP. The timing of these explosions was refined based on satellite imagery. Most tremor sources were located beneath the northeastern slope of the volcano at shallow depths. The spatiotemporal evolution of tremor activity, alongside independent volcanological data, suggest it was generated by magma-hydrothermal processes activated by and intrusion of fresh magma, which caused the hydrothermal explosions. Thus, observations of such seismic tremor may provide a short-term precursor for the future hydrothermal explosions.

We perform Rayleigh wave ambient noise tomography to investigate crustal seismic velocity structure and sources of volcanism in Armenia. Armenia, a key part of the tectonically and volcanically active Caucasus‐Anatolia region, is actively being deformed by the ongoing Arabian‐Eurasian continental collision. Unlike typical intracontinental settings, Armenia exhibits exceptional diversity of volcanic compositions and eruption styles: large stratovolcanoes are interspersed among more broadly distributed monogenetic cones and extensive lava flows. This study presents the first seismic tomography model of Armenia with sufficient resolution to infer potential magma sources. We analyze ∼19 months of continuous ambient noise data recorded by 32 seismic stations, extracting Green's functions and Rayleigh wave dispersion curves. A two‐step tomographic inversion first yields 2D group velocity maps, followed by a 3D shear‐wave velocity model. Synthetic tests confirm the model's resolution and ability to detect lateral and vertical velocity anomalies. Our results reveal prominent low‐velocity anomalies down to 25 km beneath monogenetic cones, likely indicating magma transport zones. At greater depths, velocity anomalies reverse sign. A high‐velocity zone at 40 km depth beneath dispersed cones suggests crustal thinning and asthenosphere upwelling. Beneath Lake Sevan, we identify two distinct structures: a low‐velocity anomaly in the NW linked to fault‐related fracturing and fluid saturation, and a high‐velocity anomaly in the SE that may represent a rigid block, possibly remnant oceanic crust. This study provides new insights into crustal structure beneath Armenia, shedding light on its magmatic and tectonic evolution.

The 2021 eruption of Cumbre Vieja on La Palma, Canary Islands, provided a unique opportunity to investigate the dynamics of magma migration and storage during a large effusive eruption. In this study, we employ repeated seismic tomography to image temporal changes in the subsurface structure beneath La Palma, using body‐wave travel times from local earthquakes recorded before and during the eruption. By carefully selecting paired data sets with identical numbers of events and similar ray path distributions, we minimize biases introduced by variations in seismicity patterns and ensure robust detection of velocity changes. Our results reveal a complex magma plumbing system characterized by a deep magma storage zone below 8 km depth and a shallow, fluid‐saturated region extending to ∼3 km depth. During the eruption, a high Vp/Vs anomaly, interpreted as a magma conduit, formed and ascended through a rigid barrier at 5–8 km depth, facilitating the transport of significant volumes of magma to the surface. The conduit evolved from a diapir‐like structure with a stable head at 6–7 km depth in the early stages of the eruption to an ascending plume reaching 3 km depth in later stages. This study highlights the utility of repeated seismic tomography in unraveling the dynamic processes driving effusive eruptions and provides new insights into the evolution of magma systems during volcanic unrest.

The Armenian Highlands, a tectonically active segment of the Arabia‐Eurasia collision zone, exhibit widespread Quaternary volcanism, rapid uplift, and intense seismicity. However, the lithospheric processes driving these phenomena remain poorly understood. We present the first P‐ and S‐wave body‐wave tomography model of Armenia's crust and uppermost mantle, derived from regional earthquakes recorded by both permanent and temporary seismic networks. Our model reveals that beneath the Armenian Highlands, the crust is underlain by a pronounced low‐velocity anomaly in the uppermost mantle (∼40 km depth), indicative of asthenospheric upwelling and melt generation. This anomaly spatially correlates with monogenetic volcanic fields (e.g., Gegham and Syunik), supporting magma ascent through thermally weakened conduits. The lithosphere adheres to a “crème brûlée” rheological model, characterized by a seismogenic brittle layer (10–25 km depth, marked by high‐velocity anomalies) atop a ductile lower crust and uppermost mantle (with negative Vp and Vs anomalies). Beneath the Gegham Plateau, a thinner brittle high‐velocity layer coincides with elevated seismicity at 10–20 km depth, suggesting magmatically induced earthquakes rather than purely tectonic strain release. Collectively, this work establishes a framework for understanding collision‐related magmatism in comparable orogens.

The Mutnovsky and Gorely volcanoes in Kamchatka, located 70–80 km southwest of Petropavlovsk‐Kamchatsky, pose significant hazard due to their potential for explosive eruptions. Mutnovsky also hosts the Mutnovsky Geothermal Power Plant (MGPP). This study presents the first crustal‐scale three‐dimensional seismic velocity model derived from ambient noise tomography, utilizing data from a temporary 2023–2024 seismic network of 65 mixed broadband and short‐period stations, in addition to four permanent stations. The model reveals multiple low‐velocity zones: a low‐velocity anomaly at 2–5 km depth below sea level is interpreted as a Mutnovsky magma chamber, while the other one at 2.5–5 km depth beneath the MGPP likely reflects an active magmatic intrusion sustaining the geothermal system. A shallow anomaly at 0.5–1 km depth beneath the MGPP is attributed to production intervals associated with geothermal boreholes. Furthermore, the model indicates hydrothermal connectivity between the Mutnovsky field and the Zhirovskoy Valley, with no apparent connection to the Vilyuchinsky Valley springs. Beneath the Gorely caldera, a wide low‐velocity zone suggests the presence of unconsolidated sediments and an underlying magmatic intrusion at 2–4 km depth.

The Lousal Mine (Iberian Pyrite Belt, Portugal) was operated from 1900 to 1988 for the extraction of massive sulphides and was later rehabilitated as a science museum. It was selected as a test site for underground muon tomography applied to geophysical surveys, as part of the LouMu project. This study focuses on seismic tomography to analyse the subsurface above the mine gallery, primarily surveyed by a muography telescope, which was developed specifically for this site by the Laboratory of Instrumentation and Experimental Particle Physics. To validate the muon tomography results, an initial approach using conventional 2D seismic refraction failed to reach the Waldemar gallery depth, due to limited seismic ray coverage. Therefore, an innovative setup using surface shots and in-gallery geophones was implemented, providing full ray coverage. A 3D velocity model was then produced using the ATOM3D code, which enabled the integration of this configuration and performed travel-time inversion for velocity calculation. A regional dextral strike-slip fault, the Corona Fault (CF), crosses the surveyed area, and served as the main focus of this investigation. The 3D velocity model successfully detected this structure, that corresponded to the boundary between positive anomalies of the Volcano-Sedimentary Complex (VSC) and negative anomalies of the Phyllite-Quartzite Group (PQG). The absolute velocity distribution showed a distinct offset around the Corona Fault (CF), indicating a dextral strike-slip mechanism. A subvertical extension of secondary faults was observed, reflecting deformation similar to that of the main tectonic context. Previous data from the gallery confirmed that these results are consistent with the known geology and can serve as a reference for the muon tomography interpretations.

Characterising the inner structure of a volcano is critical for volcano monitoring. Typically, the output models employ a static view of the subsurface, although the experiments can utilise signals spanning a wide temporal range. We analyse the 2020-2021 pre-, co- and post-eruptive data at La Soufrière volcano in Saint Vincent. We use different imaging techniques to extract the main features of the subsurface. The results show: a solidified primary and secondary magma cap during the effusive phase, consistent with previous geochemical studies. Moreover, changes in the 1D velocity structure as the magma intrusion, the hydrothermal system and the conduit stability interact. Finally, we observe weak ”flow-through” reservoirs that promote fragmentation at various depths, consistent with explosion depths. For the case of La Soufrière, updating the velocity structure in quasi-real-time does not improve the accuracy since the location errors are usually larger; however, in other systems/networks the applicability must be assessed.