Lucia Gurioli

Dynamics of Strombolian explosions: Inferences from field and laboratory studies of erupted bombs from Stromboli volcano

Strombolian activity is characterized by repeated, low energy, explosions and is named after the volcano where such activity has persisted for around 2000 years, i.e., Stromboli (Aeolian Islands, Italy). Stromboli represents an excellent laboratory where measurements of such explosions can be made from safe, but close, distances. During a field campaign in 2008, two 15 cm diameter bombs were quenched and collected shortly after a normal explosion. The bombs were characterized in terms of their textural, chemical, rheological, and geophysical signatures. The vesicle and crystal size distribution of the samples, coupled with the glass chemistry, enabled us to quantify variations in the degassing history and rheology of the magma resident in the shallow (i.e., in last 250 m of conduit length). The different textural facies observed in these bombs showed that fresh magma was mingled with batches of partially to completely degassed, oxidized, high-crystallinity, high-viscosity, evolved magma. This magma sat at the top of the conduit and played only a passive role in the explosive process. The fresh, microlite-poor, vesiculated batch, however, experienced a response to the explosive event, by undergoing rapid decompression. Integration of geophysical measurements with sample analyses indicates that popular bubble-bursting models may not fit this case. We suggest that the degassed, magma forms a plug, or rheological layer, at the top of the conduit, through which the fresh magma bursts. In this model we need to consider the paradox of a slug ascending too fast through a magma of varying viscosity and yield strength.

Interaction of pyroclastic density currents with human settlements: Evidence from ancient Pompeii

Integrating field observations and rock-magnetic measurements, we report how a turbulent pyroclastic density current interacted with and moved through an urban area. The data are from the most energetic, turbulent pyroclastic density current of the A.D. 79 eruption of Vesuvius, Italy, which partially destroyed the Roman city of Pompeii. Our results show that the urban fabric was able to divide the lower portion of the current into several streams that followed the city walls and the intracity roads. Vortices, revealed by upstream particle orientations and decreases in deposit temperature, formed downflow of obstacles or inside cavities. Although these perturbations affected only the lower part of the current and were localized, they could represent, in certain cases, cooler zones within which chances of human survival are increased. Our integrated field data for pyroclastic density current temperature and flow direction, collected for the first time across an urban environment, enable verification of coupled thermodynamic numerical models and their hazard simulation abilities.

Transitions between fall phases and pyroclastic density currents during the AD 79 eruption at Vesuvius: building a transient conduit model from the textural and volatile record

The magmatic phase of the AD 79 eruption of Vesuvius produced alternations of fall and pyroclastic density current (PDC) deposits. A previous investigation demonstrated that the formation of several PDCs was linked with abrupt increases in the proportion of denser juvenile clasts within the eruptive column. Under the premise that juvenile clast density is controlled by vesiculation processes within the conduit, we investigate the processes responsible for these variations at or close to fragmentation levels. Pumice textures (vesicle sizes, numbers, and connectivity combined with crystal textures) from the AD 79 PDC deposits are compared to those from interbedded fall samples. Both PDC and fall deposits preserve textures that represent a full spectrum of degassing and outgassing processes, from bubble nucleation to collapse. Combining the textural and volatile (groundmass H2O) data, we derive a conduit model that satisfies all the textural and physical observations made for this phase of the eruption: lateral vesicularity/density stratifications are produced by maturing of bubble textures with superimposed localized shearing of bubble-rich magmas, which enhance outgassing of H2O. The incorporation of denser slower-moving magma from the conduit margins (“lateral magma density gradient”) is likely to be responsible for the higher abundances of dense juvenile pumice that triggered partial column collapses. We also illustrate how variations in the fragmentation depth (tapping a “vertical magma density gradient”) can be responsible for variations in erupted clast density distributions, and potentially in the extent of degassing/outgassing.

Classification, landing distribution, and associated flight parameters for a bomb field emplaced during a single major explosion at Stromboli, Italy

We propose a novel approach to studying a ballistic bomb deposit. Favorable circumstances, a unique dispersal axis, an operational thermal video camera, and application of an innovative methodology allowed estimates of volume and mass erupted, and definition of mass partitioning between bombs of various sizes. This allowed the creation of a multidisciplinary database for a single major explosion at Stromboli volcano (Italy), the type locality of Strombolian eruptions. The dispersion and direction of the deposit were consistent with a major explosion on 21 January 2010. Field data comprised 780 mapped bomb locations and sizes, and were organized into a GIS with a lidar-derived digital elevation model as its base. This allowed us to define the landing distribution and flight parameters for erupted bombs. The data defined discontinuous deposition to build a cluster-dominated bomb field, with a total deposit volume of ∼10 m3, a mass of ∼2 × 104 kg, and a grain size dominated by large bombs (1 m in diameter). The parameters defined here for a major eruption at Stromboli show that the Strombolian style of volcanism, and its deposits, need to be treated carefully, and a different approach is needed in the future to truly characterize and classify such small (but globally common) explosive eruptions. The recognition that sedimentation from such eruptions will be uneven leads to the important conclusion that isopachs and isopleths cannot be used to estimate eruption volumes for such explosions.

Column collapse and generation of pyroclastic density currents during the A.D. 79 eruption of Vesuvius: The role of pyroclast density

The Plinian columns formed during the magmatic phase of the A.D. 79 eruption of Vesuvius alternated several times between fully stable, buoyantly rising regimes and unstable regimes of partial or total collapse. Six pyroclastic density currents (PDCs) were produced during unstable regimes, and ultimately caused the destruction of Roman towns around the volcano. Through new measurements of juvenile clast density and estimations of ascent parameters, we show that four partial collapses were likely triggered by increases in the abundance of dense juvenile clasts within the eruptive column. In contrast, the total collapse probably occurred in response to an increase in the wall-rock content injected into the plume during a progressive widening of the conduit. A sixth low-energy, small collapse resulted from high abundances in both dense juvenile clasts and wall-rock material. Simulations of eruption column behavior already account for the effects of variations in conduit radius, mass discharge rate, and particle size, but have yet to include variable clast density and wall-rock abundance that cause temporal variations in plume density. Our results suggest that both parameters can exert a significant control on the potential for generation of PDCs.

Anatomy of a Strombolian eruption: Inferences from particle data recorded with thermal video

Over the past decade, a proliferation of new technologies has pushed forward our ability to measure the dynamics of volcanic emissions as they exit, and ascend above, the vent. Measuring parameters of all particles as they exit the vent during an explosive eruption is the best way to gather parameters such as size, shape, velocity, and mass for the solid (particulate) fraction of the plume, in our case this being the lapilli and bomb component. We compute particle velocities and size distributions using high spatial resolution (centimeter-sized pixel) thermal infrared imagery collected at 200 Hz for small explosive eruptions at Stromboli (Italy). Our study covers 13 eruptions from Strombolis southwest crater that occured in October 2012, plus 13 eruptions from the southwest crater, and 5 eruptions from the northeast crater in May 2014. We obtain a statistically robust database for size, mass, and velocity of 83,000 particles. Most particles have sizes of 5 to 15 cm so that the majority of individual particle masses are below 0.4 kg. However, 4950 (6%) of the particles are heavier than 5 kg and represent 59% of the total mass erupted. We also show that the smallest particles detected have the highest velocities with the maximum recorded vent-leaving velocity being 240 m/s. While the thermal data provide insights into particle emission and launch dynamics, correlation with seismic data sheds light on the source mechanism. Our results lead us to suggest that pyroclast-dominated explosions are a consequence of the presence of a viscous, degassed cap at the head of the magma column, whereas gas-dominated events are a consequence of slug bursting in a “cleaner” conduit, the cap having been lost by convective overturn.

Precursory phenomena and destructive events related to the Late Bronze Age Minoan (Thera, Greece) and AD 79 (Vesuvius, Italy) Plinian eruptions; inferences from the stratigraphy in the archaeological areas

Abstract Volcanological studies in the Bronze Age settlement of Akrotiri (Santorini, Greece) and in the Roman towns of Pompeii and Herculaneum (Vesuvius, Italy) have provided information about the precursory phenomena preceding the Minoan and AD 79 Plinian eruptions and the impact of the eruptive products on the human settlements. The Akrotiri settlement was badly damaged by earthquakes before the onset of the eruption. A building debris layer, related to these earthquakes, covers the Minoan soil. The fallout pumice bed, mantling the ruins, freezes a state of partial destruction of the settlement. The deposition of the following pyroclastic flows completed the covering of the site. Strong seismicity also occurred during the opening and the Plinian phases. At the Herculaneum and Pompeii excavations clear evidence of strong pre- and syn-eruptive earthquakes is absent. Herculaneum, just 7 km west of the crater of Vesuvius, was destroyed by several pyroclastic flows, which buried the town under 20 m of deposits. Pompeii was covered by a 3 m thick blanket of pumice fall deposit. Distal dilute and turbulent ash clouds reached the town toward the end of the Plinian phase, killing all remaining inhabitants. The following turbulent cloud related to the onset of the caldera collapse completely destroyed the town, which was successively covered by the final phreatomagmatic products of the eruption.

Conduit Dynamics in Transitional Rhyolitic Activity Recorded by Tuffisite Vein Textures from the 2008–2009 Chaitén Eruption

The mechanisms of hazardous silicic eruptions are controlled by complex, poorly-understood conduit processes. Observations of recent Chilean rhyolite eruptions have revealed the importance of hybrid activity, involving simultaneous explosive and effusive emissions from a common vent. Such behaviour hinges upon the ability of gas to decouple from magma in the shallow conduit. Tuffisite veins are increasingly suspected to be a key facilitator of outgassing, as they repeatedly provide a transient permeable escape route for volcanic gases. Intersection of foam domains by tuffisite veins appears critical to efficient outgassing. However, knowledge is currently lacking into textural heterogeneities within shallow conduits, their relationship with tuffisite vein propagation, and the implications for fragmentation and degassing processes. Similarly, the magmatic vesiculation response to upper conduit pressure perturbations, such as those related to the slip of dense magma plugs, remains largely undefined. Here we provide a detailed characterization of an exceptionally large tuffisite vein within a rhyolitic obsidian bomb ejected during transitional explosive-effusive activity at Chaiten, Chile in May 2008. Vein textures and chemistry provide a time-integrated record of the invasion of a dense upper conduit plug by deeper fragmented magma. Quantitative textural analysis reveals diverse vesiculation histories of various juvenile clast types. Using vesicle size distributions, bubble number densities, zones of diffusive water depletion, and glass H2O concentrations, we propose a multi-step degassing/fragmentation history, spanning deep degassing to explosive bomb ejection. Rapid decompression events of ~3-4 MPa are associated with fragmentation of foam and dense magma at ~200-350 metres depth in the conduit, permitting vertical gas and pyroclast mobility over hundreds of metres. Permeable pathway occlusion in the dense conduit plug by pyroclast accumulation and sintering preceded ultimate bomb ejection, which then triggered a final bubble nucleation event. Our results highlight how the vesiculation response of magma to decompression events is highly sensitive to the local melt volatile concentration, which is strongly spatially heterogeneous. Repeated opening of pervasive tuffisite vein networks promotes this heterogeneity, allowing juxtaposition of variably volatile-rich magma fragments that are derived from a wide range of depths in the conduit. This process enables efficient but explosive removal of gas from rhyolitic

Simulating the thermorheological evolution of channel-contained lava: FLOWGO and its implementation in EXCEL

Abstract FLOWGO is a one-dimensional model that tracks the thermorheological evolution of lava flowing down a channel. The model does not spread the lava but, instead, follows a control volume as it descends a line of steepest descent centred on the channel axis. The model basis is the Jeffreys equation for Newtonian flow, modified for a Bingham fluid, and a series of heat loss equations. Adjustable relationships are used to calculate cooling, crystallization and down-channel increases in viscosity and yield strength, as well as the resultant decrease in velocity. Here we provide a guide that allows FLOWGO to be set up in Excel. In doing so, we show how the model can be executed using a slope profile derived from Google™ Earth. Model simplicity and ease of source-term input from Google™ Earth means that this exercise allows (i) easy access to the model, (ii) quick, global application and (iii) use in a teaching role. Output is tested using measurements made for the 2010 eruption of Piton de la Fournaise (La Réunion Island). The model is also set up for rapid syneruptive hazard assessment at Piton de la Fournaise, as we show using the example of the response to the June 2014 eruption.

Modern Multispectral Sensors Help Track Explosive Eruptions

Due to its massive air traffic impact, the 2010 eruption of Eyjafjallajokull was felt by millions of people and cost airlines more than U.S.