V51D-0774
Time-lapse Images of Dacite Dome Growth at Mount St. Helens, Washington, 2004–2007
Fixed-position, digital cameras of varying type have been used to record the growth and changing morphology of a dacite lava dome at Mount St. Helens since the onset of extrusion in October 2004. Dome growth involves extrusion of a solidified lava plug at rates, determined from sequential digital elevation models derived from vertical aerial photographs, that have declined from 6 m3/s in late 2004 to 0.3 m3/s in early 2007. Time- lapse images at 1-10 megapixel resolution have been acquired at rates of 1-60 frames per hour over periods of several hours to nearly three years. Images were taken at camera sites established on the crater floor, crater rim, 1980-1986 lava dome, and currently active dome. In addition to recording the physical changes of the growing dome and its dramatic effect on a surrounding glacier, the cameras have captured seismogenic rockfalls, slumps, and small explosions. Quantitative analysis of the oblique images requires epipolar equations to calculate coordinates of mobile points that are visible from two or more cameras to yield three-dimensional surface displacements that are consistent with displacements measured by other means (for example, GPS and short duration photographic sequences at known scale) and with the generally declining extrusion rate. Although extrusion has been continuous, the photographs show episodes of strong differential motion within an extruding plug and several short pauses, at least two of which had coincident occurrences of relatively large (magnitude 3) earthquakes when motion resumed. Several observations from the photographs are consistent with a proposed stick-slip model of plug motion (Iverson and others, Nature, 23 November 2006), but many observations are not at sufficient frequency to discriminate between periodic and continuous plug motion. http://vulcan.wr.usgs.gov/home.html
V51D-0775
Ground Tilt Associated with the 2004-Present Eruption of Mount St. Helens, Washington
Tiltmeters in the near-vent region of effusively erupting volcanoes may shed important light on the mechanics of lava extrusion and dome growth. Tiltmeters installed in the crater at Mount St. Helens, Washington, during the ongoing eruption (2004 to present) have recorded cyclic, repeating tilt events throughout the course of the eruption. Individual tilt events last from minutes to hours, are remarkably coherent between different tiltmeters within the crater, and frequently display a clear saw-tooth shaped temporal evolution. Azimuths of the tilt events have remained remarkably consistent over time, and point to a source located generally beneath the south part of the 1980-86 lava dome. Local seismic activity strongly affects the tilt signals, introducing both an instrument response unrelated to real ground tilt, and also in many cases triggering real tilt events that may last for many minutes. Teleseisms are easily detected by the tiltmeters, and have also been observed to trigger longer-period tilt events. It seems likely that ground tilt is related to the motion of (nearly solid) magma as it moves up the conduit at Mount St. Helens, and its diversion to the south beneath the 1980-86 lava dome. While the eruption appears to be steady over longer time scales, the tilt signals require episodic behavior. Inversion of the tilt data for a generalized source should reveal the kinematics of this transient process.
V51D-0776
Laboratory Studies of High Temperature Deformation and Fracture of Lava Domes
The high temperature fracture mechanics of magma at high temperatures exerts a fundamental control on the stability of lava domes and the timing and style of eruptions at andesitic to dacitic volcanoes. This is evidenced in the pervasive fracturing seen in both ancient and active magma conduits and lava domes; in addition to the volcanic earthquakes that occur before and during episodes of dome growth and dome collapse. Uniaxial and triaxial deformation experiments have been performed on crystal rich and crystal free magmas (andesite from Ancestral Mount Shasta, California, USA and a rhyolitic obsidian from Krafla, Iceland) at a range of temperatures (up to 900°C), confining pressures (up to 50 MPa) and strain rates (10-5s-1) to 10-3s-1) whilst recording acoustic emissions (AE). Results from these experiments provide useful inputs into models of lava dome stability, extrusion mechanisms, and source mechanisms for volcanic earthquakes. However, the large sample sizes used to ensure valid results (25mm diameter and 75mm length) made it difficult to maintain stable high temperatures under confined conditions. Also, only rudimentary AE data could be obtained, due to the distance of the transducers from the samples to keep them away from the high temperatures. Here, we present modifications to this apparatus, which include a new furnace, improved loading system, additional pore pressure and permeability measurement capability, and vastly improved acoustic monitoring. This allows (1)stable higher temperatures (up to 1000°C) to be achieved under confined conditions, (2) high temperature and moderate pressure (up to 70 MPa) hydrostatic measurements of permeability and acoustic velocities, (3) high temperature triaxial deformation under different pore fluid and pressure conditions, and (4) full waveform AE monitoring for all deformation experiments. This system can thus be used to measure the physical properties and strength of rocks under volcanic conditions and to simulate volcanic earthquakes.
V51D-0777
Non-Newtonian rheology of Unzen : crystallinity as a shear-thinning factor.
The dome building eruptions of Unzen generated repeated dome failure and pyroclastic flows. The domes varied in character and behavior from exogenic to endogenic. The deformation of highly crystallised dome lavas is a key to understanding their rheology and to fixing their failure criteria. In this study we investigate the stress and strain- rate dependence of Mt Unzen dome lavas. Their rheology has been determined for temperatures from 900 to 1010°C and stresses from 2 to 60 MPa in uniaxial compression. Two kinds of non-Newtonian behavior are observed. The first is instantaneous and on the whole recovered during stress release. The second is time- dependent and non-recoverable. These effects are termed Instantaneous and Delayed Apparent Viscosity Decrease (IAVD & DAVD), respectively. The IAVD is typical of that observed in previous experiments on crystal-bearing melts. It has also been observed for crystal-free melts at much lower magnitude. We infer that the crystal phase responds elastically to the stress applied and relaxes once the load is withdrawn. The DAVD appears more complex and this regime depends on the stress (and/or strain-rate) history. We distinguish four different domains: Newtonian, non-Newtonian, crack propagation and failure domains. Each of this domains, expresses itself as a different regime of viscosity decrease. Due to stress localization, cracking appears in crystal-bearing melts (intra-phenocryst and/or the in the melt matrix) earlier than in crystal-free melts. For low stresses, the apparent viscosity is higher for crystal-bearing melts as predicted by Einstein-Roscoe equations. However, while the stress (or strain rate) increases, the apparent viscosity is decreasing to that of the crystal-free melt and could be even lower if viscous heating effects are involved. The crystalline phase is commonly believed to increase the viscosity according to the Einstein-Roscoe equations. Indeed, those equations are confirmed here for low stresses and strain rates. However, more importantly, the presence of the crystalline phases results in an apparent viscosity that becomes strongly stress and strain-rate dependent. Einstein-Roscoe overestimates THIS apparent viscosity by several orders of magnitude. This study demonstrates the dominance of non-Newtonian rheology in understanding the extrusion of dome lavas at Mt Unzen.
V51D-0778
The Evolution of Elastic Moduli With Increasing Crack Damage During Cyclic Stressing of Etna Basalt
Volcanic edifices, such as Mt. Etna volcano, are commonly subject to cycles of pressurization and depressurization over extended periods of time due to repeated episodes of magma emplacement from deep reservoirs to shallow depths. Such repeated episodes of deformation can lead to an increase in the level of crack damage within the rocks of the edifice, and hence changes in their elastic properties. Importantly, a number of volcano monitoring techniques, such as seismic tomography and ground deformation modeling, rely on accurate knowledge of elastic properties. However, the effect of cyclic stressing on mechanical and elastic properties of volcanic rock remains unclear. To this end, we report results of changes in elastic moduli from stress-cycling experiments on samples of extrusive basalt from Mount Etna, Italy. The basalt contains an extensive pre-existing network of isotropic, interconnected microcracks caused by cooling. Both oven-dry and water-saturated samples were initially loaded to 20 MPa at a constant rate and then unloaded to 8 MPa. Samples were then sequentially reloaded and unloaded at the same rate with the peak stress in each subsequent cycle increased by 10 MPa. Stress-cycling was continued until each sample failed. Results from oven-dry samples showed a gradual reduction in sample stiffness with each increasing stress cycle that resulted in a total decrease in Young's modulus of approximately 30% and an increase in Poisson's ratio of approximately 60%. Results from water-saturated samples showed an almost identical trend. These changes in moduli are attributed to the growth of new cracks in each stress cycle and, hence, an increase in the total crack density. This is supported by the observation of increased acoustic emission (micro-seismic) output in each cycle. We also observed the Kaiser stress-memory effect, where acoustic emission on each cycle only occurs when the maximum stress in the previous cycle has been exceeded. During the deformation history of volcanic edifices, however, the stress in each pressurization cycle may not always exceed that of the previous cycle. In order to better understand this more realistic situation, we also report results from cyclic stressing experiments where the peak stress in each cycle has been randomly selected to be either higher or lower than that of the previous cycle. In this case we observed a more complex manifestation of the Kaiser effect, where AE output in any cycle was only observed when the stress in that cycle exceeded the maximum stress on any previous cycle.
V51D-0779
Acoustic emissions accompanying the compressive ductile-brittle transition in highly- crystalline lavas.
Understanding of the ductile-brittle transition in dome lavas may well contain the key to an adequate description of dome growth and stability. To elucidate this transition in dome lavas, a series of experiments were performed to characterize microcracking during compressive deformation of crystal-rich lavas. Multiphase lavas behave as visco-elastic fluids with a strain-rate dependence of viscosity across the ductile-brittle field. In order to map out the onset of brittle failure across the transition, we have deformed large volume samples (80 mm long by 40 mm diameter) in a high-load, high-temperature uniaxial press equipped with acoustic emission (AE) monitoring sensors. Our apparatus has been calibrated using an NBS717a standard glass. The absence of cracking and associated AE during deformation of this standard, which behaves as a homogeneous viscous melt under our experimental conditions, allows us to calibrate and filter out extraneous background noise. Samples from each of the five volcanoes chosen for this study (Colima, Unzen, Bezimianny, Krakatau, and Tungurahua) were deformed at two temperatures (940 and 980°C) and at stresses from 1 to 50 MPa. At low stresses (1-10 MPa), only a few AE events were detected and the AE rate decreased with increasing strain. Occasional high-energy events were recorded, and attributed to cracking of single crystals. Increasing the stress to 20-30 MPa resulted in an increased AE rate that stayed essentially constant with increasing strain. Occasional high-energy events persisted. At 40 and 50 MPa, the AE rate was higher still, and increased with increasing strain (overwhelming the few high energy events that continued to occur). Preliminary evaluation of the seismic b-value shows a decreasing trend from >3.0 at low stress to <1.5 at high stress, suggesting a shift from distributed small-scale cracking to more localized larger-scale cracking as stress is increased. These results will be discussed in terms of the deformation of dome lavas during extrusion and residence at the surface.
V51D-0780
Pre-eruptive seismicity associated with explosive events at Santiaguito volcano, Guatemala
We conducted a one-week field campaign to acquire seismo-acoustic data together with other geophysical observations (thermal, radar, video, gas flux) at Santiaguito volcano, Guatemala in January 2007. The project goal was to understand and quantify energy budgets and physical mechanisms associated with explosions and lava effusion at a silica-rich volcano. The Santiaguito volcanic dome complex in western Guatemala provided an ideal laboratory as it has displayed a complex mix of explosive and effusive activity since it began to grow in 1922. In recent years, small discrete explosive events have occurred on the order of once or twice per hour, while a contemporaneous dacitic-andesitic lava flow is extruded to the south/south-east. We are currently seeking an improved understanding of the enigmatic seismicity associated with these explosions, which comprise approximately 85% of the local earthquake catalogue. Rockfalls 10% and other events (regional tectonic, volcano-tectonic and hybrid) 5% comprise the balance of events. During the week-long explosion earthquake catalogue of 300 events (45 per day), we identified ubiquitous "precursory seismicity" of varying length duration (4 - 17 seconds), which preceded visible pyroclastic emissions and had no coincident infrasound. The onset of the comparatively large explosion earthquake (‘primary phase') occurred simultaneously with the generation of an infrasonic pulse and a release of ash and magmatic gas. Precursory seismic signals are emergent, display a spectral peak between 0.5 and 2 Hz, and possess varying degrees of energy at higher frequencies, which is both time-varying and event-dependent. Precursory seismicity typically increases in amplitude leading up to the primary phase, which is typically also peaked below 2 Hz and is classified as a long-period earthquake. Video analysis of the dome surface at the onset of explosions shows a surge, or horizontal displacement, of the dome coincident with the primary phase long period event. Initial analysis of particle motions, wave envelope and frequencies suggests fracturing of the near-vent subsurface. The following long period event suggests an initially slow then rapid vertical motion of conduit material which leads to displacement of the lava flow on the dome's surface. The associated rapid expansion of magmatic gas produces low intensity infrasound and may also contribute to the seismic wavefield.
V51D-0781
Observational Evidence for a Rapidly Moving Seismic Source at Mount St. Helens
We analyzed highly repeatable waveforms recorded at Mount St. Helens on October 9, 2004. The earthquake clones, or "drumbeats", from this time interval marked the onset of lava extrusion in the ongoing eruption. Hourly stacks of these waveforms correlate extremely well, with the correlation coefficient decreasing as the time separation between these stacks increases. We used coda wave interferometry [Snieder and Vrijlandt, 2005] to relate the correlation coefficient between hourly stacks recorded at different times to the migration of the source that excites these waveforms. Coda wave interferometry uses the changes in the seismic coda to diagnose changes in the medium, or in the excitation of these waves. We applied our analysis to different frequency bands and to different time windows within the coda. Results suggest that the source mechanism is constant over the day-long time interval studied while the Green's function is changing due to changes in the source location or physical properties of the surrounding crust. The consistency of the inferred source migration, as measured at distinct stations, argues that changing source location, instead of changes in velocity structure along the ray path, causes the observed change in the waveforms. The estimated source migration is about 55 m/day. This value is about a factor 10 larger than estimates of the vertical displacement of the plug in Mount St. Helens. This implies that if the seismic events are caused by brittle failure on the edges of the extruding plug, the seismicity moves with respect to the plug. Snieder, R. and M. Vrijlandt, M., Constraining Relative Source Locations with Coda Wave Interferometry: Theory and Application to Earthquake Doublets in the Hayward Fault, California, J. Geophys. Res., 110, B04301, 10.1029/2004JB003317, 2005.
V51D-0782
Using the Seismic Amplitude Decay of Low-Frequency Events to Constrain Magma Properties.
Low-frequency events are considered a key part of volcanic monitoring, as they are one of the few tools available that can link surface observations directly to internal volcanic processes and properties. Our model for their generation on the Soufrière Hills Volcano, Montserrat, is brittle fracturing of the magma at the conduit walls, providing the seismic trigger mechanism, followed by conduit resonance. The attenuation of seismic waves in a viscous magma is highly dependent on the properties of the attenuating material, in particular the viscous friction, controlled by the melt viscosity, gas content and diffusivity. Therefore we can use the seismicity to gain information on these magma properties. This research uses a two-dimensional viscoelastic finite-difference model, with the attenuative behaviour of the magma parameterised by an array of Standard Linear Solids. By examining the relationship between the amplitude decay of the synthetic low-frequency events, the intrinsic attenuation and the elastic parameter contrast, this research aims to link observables such as amplitude decay of the coda directly to properties such as the magma viscosity.
V51D-0783
Comparison and Occurrence of Event Types at Bezymianny Volcano, Russia and Mount St. Helens, Washington
Bezymianny volcano, Kamchatka, Russia is 50 years into an eruptive cycle, which started with a lateral blast and edifice collapse, much like Mount St. Helens in 1980. Since 1980, the occurrence of explosions and dome building at Mount St. Helens has been similar to Bezymianny volcano between 1956 and 1980. Over the past decade, Bezymianny has had twice-yearly eruptions that include plinian eruptions and continued dome building, making it an attractive target for studying such processes. Volcanoes have heterogeneous structure and thus seismic recordings of shallow (<1 km) volcanic earthquakes are subjected to large attenuation, and surface wave conversion at small distances (>1 km) from the epicenter. The result is a protracted and relatively low frequency seismic signal that is largely due to the path, with effects of the source obscured. The seismic network at Bezymianny has 8 stations within 40 km with the closest station 5 km from the active dome. In this project, we compare seismic observations of earthquakes at Bezymianny to earthquakes observed at Mount St. Helens at equivalent distances and look for events with common characteristics in the time and frequency domains. We then look at the near-field seismic data of identified events recorded at Mount St. Helens to better understand the actual character of the seismic events near the source. We use this comparison of near- and far-field seismic data to define event types at Bezymianny volcano and then look at the occurrence of those event types during different phases of activity between 2001 and 2006. The occurrence and buildup to eruption at Bezymianny is then compared to buildups in 1980 and in 2004 at Mount St. Helens. This work is important in characterizing past and future eruptive sequences at Mount St. Helens and Bezymianny volcano, and also to begin to characterize the seismicity associated with prolonged dome building.
V51D-0784
Relating stress models of magma emplacement to volcano-tectonic earthquakes
Among the various types of seismic signals linked to volcanic processes, volcano-tectonic earthquakes are probably the earliest precursors of volcanic eruptions. Understanding their relationship with magma emplacement can provide insight into the mechanisms of magma transport at depth and assist in the ultimate goal of forecasting eruptions. Volcano-tectonic events have been observed to occur on faults that experience increases in Coulomb stress changes as the result of magma intrusions. To simulate stress changes associated with magmatic injections, we test different models of volcanic sources in an elastic half-space. For each source model, we look at several aspects that influence the stress conditions of the magmatic system such as the regional tectonic setting, the effect of varying the elastic parameters of the media, the evolution of the magma with time, as well as the volume and rheology of the ascending magma.
V51D-0785
Models of Viscosity: Strengths, Weaknesses and the Challenges
Here we present a model for predicting the non-Arrhenian Newtonian viscosity of silicate melts as a function of temperature (T) and melt composition (X), including the rheologically important volatile constituents H2O and F. The model is based on >1750 measurements of viscosity on multicomponent anhydrous and volatile- rich silicate melts. The non-Arrhenian T-dependence is accounted for by the VFT equation [log η = A + B/(T(K) - C)]. The optimization assumes a common, high-T limit (A) for silicate melt viscosity and returns a value for this limit of -4.55 (e.g., log η = 10-4.6 Pa s). All compositional dependence is ascribed to the parameters B and C and is accounted for by an additional 17 model coefficients. The model has the following attributes: i) the model covers over fifteen log units of viscosity (10-1-1014 Pa s), ii) it spans most of the compositional range found in naturally-occurring volcanic rocks using 10 major and minor oxide and two volatile components (H2O, F2O-1), iii) it is computationally continuous across the entire compositional and temperature spectrum of the database, and iv) it is capable of accommodating both strong and fragile behaviour of silicate melts. Model quality is demonstrated, in part, by how well it reproduces the original observations. However, higher-level models have logical consequences that can serve as testable predictions. For example, this model is tested by how well it predicts other transport properties including glass transition temperatures (Tg) and melt fragility (m). Values of Tg (791 - 1077 K) and m (21 - 58) calculated for 58 anhydrous melts using our viscosity model are in strong agreement with the values obtained by fitting the datasets independently (789 - 1154 K and 24 - 65, respectively). We use this approach to compare our model for silicate melt viscosity against previously published models. We find our model to be more consistent with theory, to do a better job of reproducing other melt transport properties, and to work over a wider range of melt conditions (composition and temperature). Other models fail because they use a non-Arrhenian formulation, or they are calibrated for an extremely small range of melt conditions, or they predict unphysical values of viscosity in the limits, or they are over-parameterized and cannot be extrapolated beyond the original calibration dataset. Our model predicts, within inter-laboratory experimental error, viscosity and other melt properties (i.e., Tg amd m) for most of the T-X space found in natural silicate melts. Despite its strengths, there is room for improvement. Firstly, our model incorporates the effects of H2O and F, but does not account for other important volatiles, including CO2, S, Cl. Secondly, Fe is treated as a single species whereas melts contain variable proportions of ferric and ferrous iron. Thirdly, we do not model pressure effects on melt viscosity which is needed for modeling melt transport within the lithosphere. Lastly, our model is strictly empirical, in that, the components we have chosen have no explicit or independent relationship to the structure or speciation of the silicate melt. Future models may benefit from the use of a component basis that reflects melt speciation. Giordano, D. Russell, J.K. & Dingwell, D.B. (In Review, July 07) Viscosity of magmatic liquids: A model. EPSL