GMPV10.4

Linking geophysical and geochemical observables with subsurface processes is crucial to detect volcanic unrest and better anticipate eruptions. One of the most important observables to monitor pre-eruptive volcanic activity is tremor, a more or less persistent, highly periodic, ground vibration recorded near active vents. Tremor is commonly being monitored in near real-time by volcano observatories to anticipate unrest, as it may emerge, or change properties, when subsurface pressure varies. For example, it has been observed that the dominant frequency of tremor may glide towards higher or lower values before eruptions; overtones may appear or disappear; and seismic amplitude may increase or decrease. However, similar variations can be also observed during quiescence and when activity decreases. This leads to the following questions: How does tremor actually reflect the overpressure of the subsurface? Can we infer when and where the pressure beneath active vents increases or decreases by monitoring volcanic tremor? In this work, we present a new data assimilation technique that combines new physics-based models of volcanic tremor with a machine learning-based inversion algorithm to track pressure changes beneath volcanic craters in near-real time. In particular, our inversion algorithm is based on a supervised random forest classifier trained with synthetic data, whereas our physics-based model extends from Girona et al. (2019) and is based on a stop-and-go mechanism, i.e., tremor is assumed to emerge when: (i) gas is supplied randomly to shallow levels of the volcanic plumbing system; (ii) accumulates temporarily beneath permeable caps (e.g., beneath a dome or in a leaky fracture); and (iii) transfers via permeable flow to the surface. Using this machine learning-based data assimilation technique, we find that the recent 2013 unrest phase of Kawah-Ijen volcano (Indonesia) was driven by a pressure increase in the subsurface of a factor 2-to-5. This technique is currently also being applied to unveil the pressure history of the shallow vents of Pavlof and Veniaminof volcanoes (Alaska). 

Girona, T., C. Caudron, C. Huber (2019). Origin of shallow volcanic tremor: the dynamics of gas pockets trapped beneath thin permeable media. J. Geophys. Res., doi: 10.1029/2019JB017482.

How to cite: Girona, T. and Caudron, C.: A new machine learning-based data assimilation technique to detect volcanic unrest from tremor, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7109, https://doi.org/10.5194/egusphere-egu22-7109, 2022.

The transport of magma through the crust may sometimes result in volcanic eruptions at the surface, feeding a central conduit, opening fissures on the volcano flank, or at new locations in a volcanic field. Magma travels in the brittle crust by opening its way through the surrounding rock. In addition to the fracturation of the medium, the process of diking is also controlled by the magma flow, fluid-gas phase transitions, and the heat exchange. Representing the propagation of magmatic intrusions using analog and numerical model is essential to understand the physical processes occurring in nature and to mitigate the volcanic hazard linked to the emplacement of magmatic intrusions.

In this context, we performed analog experiments of air and silicon oil injections in a solidified gelatin block. Using three cameras, we monitored the propagation of the oil-filled cracks from the front, side and top views of the tank. The processing of time lapsed pictures enables to access the crack shape (dimension and orientation), trajectory and velocity. This analog modeling technique is routinely used to simulate magmatic dike propagation in the crust. Then, taking advantage of these well constrained experiments, we could validate a novel 2D boundary element model for crack propagation coupling brittle-elastic and fluid-dynamic equations. To do so, we initiate the input and boundary conditions of our numerical simulations, using gelatin and oil parameters from the analog experiments. The outputs of the model include the crack shape, trajectory, and velocity, that is computed according to an energy conservation equation, under the assumption that fluid viscous forces are limiting the crack propagation velocity. Numerical simulations are faced with the observations from our air and oil-filled crack propagation experiments. Eventually, we applied the numerical model to the 1998 magmatic intrusion at Piton de la Fournaise volcano (La Réunion Island), confronting the timing of the of the propagation with the migration of volcano-tectonic events.

How to cite: Furst, S., Pinel, V., and Maccaferri, F.: The analog model, the numerical model and the Piton de la Fournaise : tale of a propagating dike, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-3045, https://doi.org/10.5194/egusphere-egu22-3045, 2022.

Lava flows are recurring and widespread hazards that affect areas around active volcanoes, having the potential to cause significant social and economic loss. In the last decades, physics-based models of lava flows have been proven effective and powerful tools to forecast and assess the hazard posed by effusive events. These models require different input parameters, such as the physical properties of the fluid (e.g., melt compositions, water content, rheological law, thermal properties) and the topography of the terrain. A critical parameter in physical-mathematical modelling is the effusion rate, i.e. the rate at which lava is discharged. Lava effusion rate is variable in time, strongly controlling the emplacement and run-out distance of lava flows. Nevertheless, both for assessing long-term hazards and for monitoring efforts during on-going eruptions, effusion rate is assumed to be constant or to have a bell-shaped time-dependent behavior. Here we present an analysis of the time-averaged discharge rates (TADRs, i.e. the effusion rate averaged over given periods) estimated for recent flank eruptions at Mt. Etna volcano (Italy) in order to define a possible generalized effusion rate trend to be used for the physical modeling of lava flows. The temporal series of TADRs, derived from field measurements and satellite thermal imagery, were normalized in order to obtain homogeneous curves in duration and sampling times, reducing redundancies and improving data consistency. Our analysis indicates that most of the effusion rate curves for flank eruptions of Etna are characterized by a fast waxing phase with the peak occurring between the 0.5 and 29% of the total eruption time, followed by a progressive decrease in the waning phase. By using the median values associated to the occurrence of effusion peaks and to the slope variations of descending curves in the waning phase, we estimated an averaged curve that was used to run numerical tests by means of the physics-based GPUFLOW model. Different tests were performed considering how the “characteristic effusion rate curve” could impact single vent scenarios, as well as on short- and long-term hazard maps. Statistics on the final emplacements revealed variations up to 20%, confirming the key role of the effusion rate in controlling the development of lava flow fields.

How to cite: Zuccarello, F.: On the impact of the effusion rate trend for the assessment of lava flow hazards, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-484, https://doi.org/10.5194/egusphere-egu22-484, 2022.

Etna volcano has four summit craters that are characterized by periodic strombolian and lava fountaining episodes, often associated with lava flows. In the last years, the most active was the South East Crater that on 2021 produced more than fifty paroxysms that gave rise to lava flows rapidly propagating towards East, South, and South-West. Etna summit area is visited by thousands of tourists, especially in the summertime, thus it is important to evaluate the hazard related to lava flow emplacement. For this reason, we were urged to timely map the lava flows emplaced during each paroxysm whose frequency was as high as two events in 24 hours. This task has been accomplished through the integration of different remote sensing techniques, based on data availability and weather conditions. Several satellite images (Sentinel-2 MSI, Aster, Ecostress, Skysat, Landsat-8 OLI and TIRS) allowed us to map the lava flow field at spatial resolutions from 0.7 to 90 meters. Unoccupied Aerial System (UAS) surveys also allowed to acquire visible and thermal images, with high-spatial resolution, of the lava flows. Finally, thermal images acquired from the permanent network of cameras, managed by the Istituto Nazionale di Geofisica e Vulcanologia, were re-projected into the topography at 5-meter spatial resolution. The various remote sensing data enable the mapping of the lava flows and compiling a geodatabase that registers the main geometrical parameters (e.g. length, area, average thickness). The joint exploitation of remote-sensing data acquired through multi-sensors enabled, for the first time on Etna, to timely and accurately characterize frequently occurred effusive events.

How to cite: Proietti, C., Cantarero, M., and De Beni, E.: Timely mapping and quantification of the 2021 Etna lava flows through the exploitation of multi-sensors remote-sensing data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9907, https://doi.org/10.5194/egusphere-egu22-9907, 2022.

Modelling and simulation of geophysical flows are crucial to the forecasting of the propagation extent and the assessment of the related hazards. Here we introduce a new physics-based model called GPUFLOW, which was born from our twenty years of experience in Fluid Dynamics (CFD) modelling of geophysical flows. GPUFLOW features an improved physical model for the thermal and rheological evolution of lava flows, support for debris flows without thermal dependency and a parallel implementation on graphic processing units (GPUs). We estimate the influence that the GPUFLOW input parameters have on flow emplacement through different synthetic test cases and demonstrate its reliability through the 2014 pyroclastic flow and the 2018 eruption occurred at Etna volcano. This work was supported by the INGV project Pianeta Dinamico funded by MIUR (“Fondo finalizzato al rilancio degli investimenti delle amministrazioni centrali dello Stato e allo sviluppo del Paese,” legge 145/2018), Tema 8 – PANACEA 2021.

How to cite: Bilotta, G., Ganci, G., and Cappello, A.: Modelling of geophysical flows through GPUFLOW, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10168, https://doi.org/10.5194/egusphere-egu22-10168, 2022.

Mt. Etna, in Italy, is one of the most active volcanoes in the world, whose explosive eruptions represent a serious threat to the nearby populations and producing various dangerous effects mainly on properties, crops and transports. During explosive eruptions, the real-time estimation of the mass eruption rate (MER) is challenging although crucial to mitigate the impact due to the erupted tephra. Microwave radar techniques at L- and/or X-bands, as well as thermal infrared imagery, can provide a reliable MER estimation in real-time. Using the Etna lava fountains of 3–5 December 2015 as test cases, we investigate the differences among different approaches to estimate the MER: i) the mass continuity approach (MCA); ii) the top plume approach (TPA); and iii) the surface flux approach (SFA). We also introduce a new approach, called the near source approach (NSA) that is based on the X-band radar data alone. Finally, we extend the volcanic advanced radar retrieval methodology to estimate for the first time the gas-tephra mixture density near the volcanic crater. The analysis allows us to identify the optimal real-time MER retrieval strategy, showing the potential and limitations of each method. We show that the MCA method, entirely based on the X-band radar data processing, is the best strategy with a percentage uncertainty in the MER estimation of 22.3%, whereas other approaches exhibit a higher uncertainty (26.4% for NSA, 30% for TPA, and 31.6% for SFA). We investigate and optimize the different strategies for the volume eruption rate (VER), total erupted mass and volume estimations (TEM and TEV, respectively) including their uncertainties. The MER retrieval methods, described and applied in this work, showed promising results that can be exploited to improve the tephra dispersal and fallout forecasts at Etna in near real-time. Further work might be devoted to explore new techniques, using low-cost sensors for the MER estimation and employing microwave radars as validation tools.

How to cite: Mereu, L., Scollo, S., Bonadonna, C., Donnadieu, F., Freret Lorgeril, V., and Marzano, F. S.: Ground-based remote sensing and uncertainty analysis of the mass eruption rate associated with the 3-5 December 2015 paroxysm of Mt.Etna, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-9221, https://doi.org/10.5194/egusphere-egu22-9221, 2022.

The 2021 Nyiragongo (DR Congo) eruption started on 22 May 2021, nineteen years after the last effusive eruption of 2002. The lava flows erupted from three vents, one East of the summit area and two along the southern slope, and produced two lava flows, the western of which inundated part of the Goma city, causing serious damages to population, buildings and infrastructures and stopped only at ~1 km from the Goma international Airport. Here we process a variety of satellite imagery, including visible, infrared and radar data, mapping the pre-eruptive phase, the evolution of the eruption and the post-eruptive phase. Most of the remote sensing data were acquired in the framework of Virunga Geohazards Supersite, which is part of the GEO-GSNL (Geohazard Supersite and National Laboratories) initiative. In particular we analysed: (i) Sentinel 1 (European Space Agency, ESA) and (ii) COSMOSkymed, CSK (Italian Space Agency, ASI) data providing displacement time-series and eruptive source model; (iii) Visible Infrared Imaging Radiometer Suite, VIIRS ( NASA/NOAA Suomi National Polar-orbiting Partnership) data at 375 m spatial resolution to provide thermal maps; (iv) different Pleiades (AIRBUS) triplets, at 0.5 m spatial resolution, to update the topography of the volcano.

Pre-eruptive 2020-2021 InSAR (Interferometric Synthetic Aperture Radar) analysis from Sentinel-1 and high resolution CSK data show a deflation of the summit area of Nyiragongo and Nyamuragira volcanoes amounting to few cm/yr Line of Sight (LOS). The syn-eruptive InSAR data evidence surface deformation of 70 cm LOS located South of Nyiragongo, in a wide area including the city of Goma and Lake Kivu. Modelling of the InSAR syn-eruptive data show a sub-vertical dike located from South Nyiragongo reaching Lake Kivu. The top depth is 1.5 km from the surface, and the volume variation is slightly less than 0.2 km3. Post-eruptive Sentinel-1 and CSK data showed deflation of the summit area of Nyiragongo, negative LOS surface deformation at Goma and lava cooling.

VIIRS data allowed us to see an increase in the size and temperature of the lava lake a few months before the eruption, and provided a first image of the erupted lava flow on 22 May 2021 at 22:47 GMT. Thanks to Pleiades imagery we could retrieve the lava flow area and by using a pre-eruptive topography we also provided an estimation of the erupted volumes.

Results highlight how the synergic use of multi-source, multi-temporal satellite imagery, along with innovative and automatic processing techniques, may be adopted for real-time hazard estimates in an operational environment especially in remote volcanoes with limited terrestrial networks.

This contribution is supported by the GEO-GSNL initiative and the H2020 Reliance project (grant agreement 101017501).

How to cite: Balagizi, C., Ganci, G., Trasatti, E., Tolomei, C., and Beccaro, L.: The 2021 Nyiragongo (DR Congo) eruptive crisis monitored by multi-sensor satellite remote sensing data, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12586, https://doi.org/10.5194/egusphere-egu22-12586, 2022.