V23E-01 INVITED
Insights Into Subaqueous Explosive Eruptions From the Dynamics of Underwater Explosions
During an underwater chemical explosion, the explosive is transformed into a high-pressure mixture of gases and particulate debris that expands and displaces the surroundings. This expanding mixture forms a bubble, albeit one a great deal larger than the bubbles of everyday experience. Interaction of the explosion bubble with the water surface may send water and particulate debris to great height and in various directions. Assuming that an explosion bubble is generated during a subaqueous eruption and that the bubble behaves in the same way as one generated by a high explosive, characteristics of the subaerial "splash" may be used to make quantitative inferences about the energetics of explosive subaqueous eruptions . The comparison relies on empirical relationships between two characteristics of the splash (maximum height and initial speed) and explosion depth, all cast in a suitable nondimensional form derived from the theory of underwater explosion bubbles. Available data for subaqueous basaltic eruptions can be brought into agreement with the experimental relationship for underwater chemical explosions only if the enthalpy release during the volcanic explosions is at most a few percent of the enthalpy (relative to the boiling point of water) of the erupted mass of magma. Comparable "inefficiency" has been found in a range of experiments on molten fuel/coolant interactions, which are commonly thought to be the fundamental driving mechanism of phreatomagmatic explosions. Experimental studies of underwater chemical explosions also provide a framework for understanding some sedimentological aspects of explosive subaqueous eruptions, such as the way that basaltic pyroclasts interact with liquid water after magma fragmentation and the presence of clasts that seem to have been deposited subaqueously in a dry state.
V23E-02
Textural signatures in pyroclasts of the 1875 eruption of Askja volcano: understanding conduit and vent processes in wet phreatoplinian and dry subplinian/Plinian phases.
The eruption of Askja on March 28-29, 1875 is one of four "type" phreatoplinian examples described by Self and Sparks, 1978 and the only example with documented historical records. The phreatoplinian phase was preceded and followed by purely "dry"or magmatic fragmentation during subplinian and Plinian phases, defining a sequence of fluctuating eruption intensity and contrasting style. In this study we will assess the roles of magmatic volatiles and/or external water driving phreatoplinian eruptions and the factors which control abrupt shifts between "dry" (subplinian, Plinian) and "wet" (phreatoplinian) eruption processes. Macroscopically similar populations of highly vesicular pumice characterize all three phases. Density data also shows that pumices from each fall phase have a very similar narrow range of density. The similarity between the bulk density data for all units is thus suggestive of similar vesiculation histories. We present quantitative evidence from image anlayis of selected clasts from each layer to contrast the patterns of nucleation, growth and coalescence of vesicles in melt of the phreatoplinian and plinian/subplinian pumices. Preliminary data supports two hypotheses: (1) Vesiculation was the prime driving mechanism for even the phreatoplinian phase of the 1875 eruption and (2) shifts between "dry" and "wet" phases of the 1875 eruption were not driven by conduit processes (e.g. changing ascent rate, variable amounts of vesiculation and degassing) but instead were an artifact of changing vent position and the availability of surface water.
V23E-03
Transient multiphase processes during the explosive eruption of basalt through a geothermal borehole (Namafjall, Iceland, 1977) and implications for natural volcanic flows
Experimental and numerical studies have shown that vertical flows of gas-particle mixtures are characterized by transient behavior, with development of waves of high particle concentration separated by regions of relatively clean gas. In contrast, most models of explosive flow in volcanic conduits either treat the multiphase mixture as a single fluid (pseudo-fluid approximation) and/or assume steady flow, thereby eliminating the potential for time- dependent effects related to multiphase dynamics. The 8 September 1977 explosive eruption of basaltic tephra through a geothermal borehole (Namafjall, Iceland) provides a unique test case for multiphase volcanic processes, given that its vertical extent (~1 km) is similar to that of natural volcanic conduits and its geometry is exactly known. We model this eruption by solving separate, time-dependent governing equations for conservation of mass, momentum, and energy of the gas and particle phases, allowing for drag and heat transfer between the phases. Model results are consistent with the development of transient waves of high particle concentration that propagate up the borehole, resulting in complex compressible flow phenomena along with ejection of particles in pulses in a manner that is consistent with observations at Námafjall. These transient processes occur even though the influx of gas and particles at the base of the borehole is treated as constant. Our results indicate that transient multiphase behavior is likely to be common in volcanic conduit flows, and that a key topic of future research is quantifying the types of time-dependent behaviors and their impacts on eruption column dynamics.
V23E-04
Evidence for an Unusually Energetic Basaltic Phreatomagmatic Eruption at the Table Rock Complex in South-central Oregon (USA): Using field evidence to constrain surge flow dynamics
We estimate the velocity of pyroclastic base-surges associated with the TRC2 tuff ring, one of several basaltic hydrovolcanoes located at the Table Rock Complex, Christmas Valley Basin, south-central Oregon (TRC; Pliocene-Pleistocene; first documented by Heiken, 1971, J. Geophy Res:76:5615-5626). TRC2 represents the last and most energetic eruption of the complex. Surge deposits from this eruption form a hummocky topography, which is morphologically similar to hummocks produced in storm-surge marine environments, but much larger in scale. Wavelengths range from 20-800 m perpendicular to the flow, and 20-200 m parallel to the direction of flow depending on distance from source, producing three-dimensional bowl features distributed radially around the vent. The scale of these dune forms are more than an order of magnitude larger than most measured at other hydromagmatic edifices, and instead are consistent with bedforms associated with larger scale eruptions such as the eruption of El Chichon in 1982, and the Mt St Helens lateral blast in 1980. The hummocks are composed of fine-grained, laterally continuous, centimeter to decimeter thick strata of matrix supported tuff and lapilli tuff. Individual dune hummocks truncate underlying hummocks, suggesting that the deposits are a consequence of multiple, highly erosive, dilute density currents. The most distal exposures are located 4.7 km from the vent where the hummocks are 80 m in wavelength parallel to the direction of flow and the total deposit is 4-8 m thick, suggesting the surges extended further from source but deposits have subsequently been eroded away. We used dune wavelengths and two additional features, one a ramp-up feature 4.7 km from source, and the other a 13 m high chute and pool feature 1.6 km from source, to estimate surge velocities at various distances from the vent. Results indicate initial velocities between 110-280 m s-1, which are in the range of values estimated for the Mt St Helens blast (i.e., Kieffer, 1981, Nature: 291:5816:568-570). The estimated minimum velocities at 1.6 and 4.7 km from source are 33 and 20 m s-1 respectively. These estimates are much higher than rare eye- witness measurements for base surges, which generally range from 60 m s-1 at the source, and decay to 20 m s-1 at distances of 1.5 km (i.e. Rohrer, 1965, Lawrence Radiation lab, PNE-217P:62; Moore, 1967, BullVolc.30: 337-363). Using the initial flow velocities and assuming that initial velocities resulted only from a conversion of potential energy to kinetic energy, we calculated the corresponding column collapse height. Collapse estimates range from 2500 m, corresponding to an initial velocity of 110 m s-1, and up to16000 m, corresponding to an initial velocity of 280 m s-1. As most phreatomagmatic eruption columns reach maximum heights between 1000-4000 m (i.e. Yamamoto, 2001, Geological Survey of Japan: 52:4-5:231-239; Moore, 1967), our estimates suggest that the collapse height may have been unusually high for this style of eruption. This thus expands the envelope on the energy and scale in this style of explosive volcanism.
V23E-05
High Mobility of Erosive Granular Flows and Surge Generation
Dense granular flows driven by gravity, such as avalanches or some pyroclastic flows, are active processes that participate in the evolution of the surface of the Earth and other telluric planets. They also represent natural hazards that are a threat to many populations and infrastructures. However, natural granular flows are still poorly understood. In particular, the origin of the high mobility of pyroclastic flows or the occurrence of surges during the propagation of gravitational flows along the slope are challenging questions. Fundamentally different mechanisms were employed to explain the high mobility of these flows. Among these processes, the erosion of material already present on the underlying solid topography is expected to play a significant role in the mobility of debris or pyroclastic flows and overall dynamics of transportation. When static granular material is entrained into motion by the flowing material, no-flow and flowing zones not only coexist but exchange mass and momentum. The effect of entrainment on granular flows dynamics is a key issue in volcanology where natural flows often propagate on slopes covered by the deposits of former events. The static/flowing transition is taken explicitly into account in the new theory proposed by Aranson and Tsimring [2002]. A model based on this approach is used here to simulate the spreading of granular material over an erodible bed. Numerical results show that the presence of even a very thin layer of granular material lying on the solid bed strongly increases the mobility of granular flows. Furthermore, as the thickness of the granular layer increases, the dynamics of the flowing mass drastically changes from a decelerating flow to a steadily travelling wave. Although natural gravitational flows are more complex, these results show that erosion processes could significantly increase the mobility of pyroclastic flows and make it possible to generate surges.
V23E-06
The in-situ production of ash in pyroclastic flows
Abrasion and fragmentation of pumice clasts during the propagation of pyroclastic flows has long been recognized as a potential source for the enhanced production of volcanic ash, however its relative importance has eluded quantification (Walker, 1981). The amount of ash produced in-situ can potentially affect runout distance, deposit sorting, the volume of ash introduced in the upper atmosphere, and internal pore pressure. We conduct a series of laboratory experiments on the collisional production of ash that may occur during different regimes of pyroclastic flow transport. We further parameterize the experiments of Cagnoli and Manga (2004) to determine the rate of production of frictional ash. We find that the energy of these interactions is insufficient to create a fractal particle size distribution; rather a bimodal suite of large particles and 10-100 micron ash particles are typically produced Using these laboratory experiments we can develop a subgrid model for ash production that can be included in analytical and multiphase numerical procedures to estimate the total volume of ash produced during transport. We examine numerically a range of initial flow energies and bed slopes over which the flows propagate. To simplify the problem we consider flows starting with 1 cm pumice clasts that can be broken up into 100 micron ash. We find that for most flow conditions10-20% of the initial 1 cm clasts comminutes into ash with the percentage increasing as a function of initial flow energy. Most of the ash is produced in the high-energy regions near the flow inlet, although flow acceleration on steep slopes can produce ash far from the vent. Ash produced at the frictional base of the flow and in the collisional upper regions of the flow can be redistributed through the entirety of the flow, although frictionally produced ash accumulates preferentially near its source in the bed-load. As slope increases, the relative proportion of ash generated by friction increases relative to collisional ash production in the upper regions of the flow. For most flow conditions the ash produced causes the flows to travel further and also increases the pore pressure in the flows.
V23E-07
Numerical Model of Large Bubble Bursts at Erebus Volcano, Antarctica
Erebus volcano in Antarctica has been explosively degassing from its active phonolitic lava lake for decades. For at least the past three decades a lava lake has been present in the northeast corner of the inner crater. The lava lake continually convects and large bubbles burst at the surface 2-6 events per day. A geophysical network consisting a video coverage, acoustic sensors and broadband seismometers have been monitoring activity at the lava lake for years. There is remarkable correlation between bubble burst captured on video and infrasonic signals (with maximum excess pressure as much as 100 Pa) recorded on pressure sensors located within 300 m to 800 m from the active lake. Studies of 358 lava lake infrasound sources recorded over a 98 day period in 2006 suggest that bubble rupture location on the surface is distributed over a 40 m by 50 m area of the lava lake. Many of these bubbles are at least 10 m in diameter at rupture. The size and pressurization of the bubble bursts govern the force imparted to the atmosphere and therefore the characteristic features of the infrasonic pressure transients. A series of numerical simulations of bubble bursts from the lava lake at Erebus are performed in which bubble volumes, internal pressure and lava lake temperature are variables. The simulations are conducted with the computer code Sage that is a multi-material, finite-difference code with an adaptive mesh algorithm. Conductive heat transfer (power law temperature dependence), and strength model for wall rock and lava lake crust are considered. The locations of bubble bursts are constrained from mapped locations of bubble bursts recorded in 2006. We present results from these calculations and demonstrate the effects of bubble volume (5-10 m radius), internal pressure (0.5-10 MPa), lava lake temperature (800-1000C) and source location on the infrasonic signal.
V23E-08
GPS Detection, Modeling and Energy Estimation of the Ionospheric Wave following the 2003 Explosion of the Soufriere Hills Volcano, Montserrat
Volcanic explosions or shallow earthquakes are known to trigger acoustic and gravity waves that propagate at infrasonic speeds in the atmosphere. At ionospheric heights, coupling between neutral particles and free electrons induces variations of the electron density that are detectable using dual-frequency GPS measurements. In this study, we use GPS data collected at continuous stations in the Caribbean to detect and quantify the ionospheric perturbation triggered by the explosion of the Soufriére Hills Volcano on July 13th, 2003 (Montserrat island, Lesser Antilles). We find the disturbance in an area of maximum alignment between the direction of theoretical neutral particle motion (from ray tracing) and the Earth's magnetic field, located northwest of the island, as predicted theoretically. Its frequency content shows peaks at 1 mHz and 4 mHz indicating both a gravity wave and an acoustic component, consistent with previous observations and theoretical considerations. We retrieve a horizontal velocity component of 338 m/s for the acoustic component which, given the sound speed at the maximum electron density height, implies upward propagation at an elevation angle of 63°. We model the acoustic component of the perturbation as resulting from an explosive source at ground level (N-wave pulse). We use a ray tracing technique to propagate the neutral pressure wave in the atmosphere, accounting for the dispersive characteristics of the atmosphere while conserving total acoustic energy. We couple the neutral disturbance to electron flow by integrating the continuity equation for the charge density. We then derive synthetic integrated electron content values comparable to the GPS measurements by integrating over all rays crossing satellite-to-receiver line-of-sights, accounting for the satellite displacements. Synthetic ionospheric waveforms match the observed ones well. We minimize the misfit between observed and model waveforms to estimate the total acoustic energy of the Soufriére Hills explosion.