Activity, Unrest, and Hazard Evaluation at Stratovolcanoes and Calderas IV
Presiding: G S Mattioli, University of Arkansas; M Poland, U.S. Geological Survey Cascades Volcano Observatory
V34A-01 15:30h
Surprisingly little distal ground deformation associated with the 2004-2005 eruption of Mount St. Helens, WA
The 2004-2005 eruption of Mount St. Helens, Washington, began with a small earthquake swarm on September 23, 2004. Activity peaked on September 24, gradually declined throughout the day, then increased dramatically before culminating in a series of phreatic explosions on October 1-5. By October 11, seismicity had dropped to 2-3 small (0.5-1.5 magnitude) earthquakes per minute with larger events (2-3 magnitude) occurring at a rate of several per day. This decrease coincided with the onset of lava dome extrusion, which has continued without pause through early February 2005. Deformation of the volcano and its surroundings has been monitored primarily by a combination of GPS and InSAR. Only one dual-frequency and one single-frequency GPS unit were running at the onset of activity. On September 26, 2004, several campaign GPS stations on and around the volcano were reoccupied, and by mid-October a network of 10 continuous dual-frequency stations had been established (installed by both the USGS and UNAVCO Inc.- Plate Boundary Observatory). Measurements from locations outside the crater show surprisingly little deformation. A site 8 km north of the volcano abruptly moved 1 cm towards the crater between about September 23 and October 5, suggesting volume loss at depth, and has moved an additional 1 cm through early February. Unfortunately, these data cannot be corroborated since no other stations were operating at this distance during the period of most rapid displacement. A second dual-frequency site high on the SE flank of the volcano moved 4 cm SE between mid-November 2004 and January 2005, likely due to impingement of the new lava dome against the SE crater wall. A systematic deformation pattern is not readily apparent from analysis of GPS data from other sites. Similarly, InSAR results have shown no significant surface displacements before or during the eruption. Inside the crater, surface motion is monitored by a network of single-frequency GPS receivers that have been placed by helicopter onto both the 1980-1986 and 2004-2005 lava domes. Results suggest that the 1980-1986 dome has been moving north, away from the site of the current eruption, at rates of 0.5-1 cm/day. GPS units on the active dome have recorded displacement rates as high as 10 m/day. Together with time-lapse photography, these measurements demonstrate that the velocity of lava extrusion remained nearly constant between November 2004 and February 2005. Results from the mid- and far-field GPS stations do not agree with classic models of volcano deformation that predict pre-eruption inflation and co-eruption deflation; thus, monitoring of surface displacements has largely failed to indicate the course of the eruption. This suggests a need to reexamine models of volcano deformation so that future monitoring efforts at Mount St. Helens and other volcanoes are more effective at mapping subsurface magma migration.
V34A-02 15:45h
Inferences on Volcano Behaviour and Erupted Volume Based on Times Series of Surface Deformation: Limitations and Physical Models.
New spaced based deformation data have greatly improved the monitoring of volcanoes. Times series of surface deformation that span different eruptive episodes are now available for several stratovolcanoes and show cycles of inflation induced by magma accumulation at small depth (<10 km). Shallow storage zones are fed by replenishment from deeper sources and deformation data provide an insight inside the plumbing system. However in term of risk management,a direct comparison between the level of surface deformation currently observed and the one recorded just before the preceding eruption is not a sufficient information to make any prediction for the timing and amplitude of the next event to occur. One reason for this, is that conditions for rock failure and eruption initiation may change drastically from one eruptive cycle to another mainly because of the mass repartition changes induced by the eruption itself. The eruptive products contribute to the building of the edifice: a few millions m3 dome may rise in several months. Stress changes induced by mass redistribution can even be more important in case the edifice gets partially destroyed during the eruptive event: several km3 were removed in a few seconds on May 1980 at Mount St Helens. Such load variations around a volcanic edifice act both to induce a surface deformation signal and to produce pressure changes inside and around the storage zone. These pressure changes act in turn to change failure conditions and exsolved gas content. Variation in exsolved gas content modifies magma compressibility which may induce variations in the time of quiescent between two consecutive eruptions and the volume of eruptive products. We present some analytical and numerical models for pressure changes and surface deformation induced by magma chamber replenishment and withdrawal during an eruptive cycle. We then estimate perturbations induced by a sudden partial destruction of the volcanic edifice. Different volatiles contents are investigated. We show how these calculations can be used as a guideline to interpret deformation data in order to constrain both the geometry and pressure conditions of the plumbing system and improve hazard assessment.
V34A-03 16:00h
Diverse deformation patterns of Aleutian stratovolcanoes from InSAR imaging
Interferometric synthetic aperture radar (InSAR) is capable of measuring ground-surface deformation with centimeter to subcentimeter precision with a spatial resolution of tens-of-meters over a relatively large region. With its global coverage and all-weather imaging capability, InSAR is an important measurement technique for constraining magma dynamics of volcanoes over remote regions such as the Aleutian Islands. The spatial distribution of surface deformation data, derived from InSAR images, enables the construction of detailed mechanical models to enhance the study of magmatic and tectonic processes. This paper highlights the results from the InSAR-based survey of ground surface deformation of stratovolcanoes in the Aleutian. These results include: 1) inflation of Mount Peulik volcano preceding a seismic swarm that occurred in 1998; 2) persistent volcano-wide subsidence at Aniakchak volcano; 3) magmatic intrusion and the associated tectonic stress release at Akutan volcano; 4) magmatic intrusion at Makushin volcano associated with a small eruption in 1995; 5) complex patterns of transient deformation during and after the 1992-1993 eruption at Seguam volcano; 6) surface subsidence caused by decrease in pore fluid pressure in an active hydrothermal system beneath Kiska volcano; and 7) a lack of expected deformation associated with recent eruptions at Shishaldin, Pavlof, Cleveland and Korovin volcanoes. These studies demonstrate that the deformation patterns and the associated magma supply mechanisms over stratovolcanoes in the Aleutian are diverse and vary between volcanoes, and therefore have significant implications for future volcano monitoring effort.
http://edc.usgs.gov/Geo_Apps/
V34A-04 16:15h
Understanding and modeling volcanotectonic processes that generate surface deformation on active stratovolcanoes
Surface deformation on stratovolcanoes is the result of local stresses generated by various volcanotectonic processes. These processes include changes in fluid pressure in the associated geothermal fields and magma chambers, regional seismic or tectonic events, fault development, and dike injections. Here the focus is on magma-chamber pressure changes and dike injections. Surface deformation associated with magma-chamber pressure changes is normally referred to as inflation when the pressure increases, and as deflation when the pressure decreases. The processes that lead to inflation are primarily addition of new magma to the chamber and rapid exsolution of gas from the magma in the chamber. The processes that lead to deflation are primarily cooling (and contraction) of magma in the chamber, regional tectonic extension of the crust holding the chamber, and eruption and/or dike injection. Injection of dikes (including inclined sheets) is common in most active stratovolcanoes. However, no dike-fed eruptions can take place unless the local stress field within the volcano is favorable to feeder-dike formation. By contrast, if at any location - in any layer - in the stratovolcano the stress field is unfavorable to dike propagation, the dike becomes arrested and no eruption occurs. Detailed studies of dikes in stratovolcanoes worldwide indicate that most dikes become arrested and never reach the surface. However, arrested dikes may give rise to surface deformation, such as is commonly monitored during volcanic unrest periods. By definition, stratovolcanoes are composed of numerous alternating strata (layers) of pyroclastic material and lava flows. Commonly, these layers have widely different mechanical properties. In particular, some layers such as lava flows and welded pyroclastic flows may be stiff (with a high Young's modulus), whereas other layers, such as non-welded pyroclastic units, may be soft (with a low Young's modulus). Here I present new numerical models on the surface deformation on typical stratovolcanoes. The models show, first, that the surface deformation during magma-chamber inflation and deflation depends much on the chamber geometry, the loading conditions, and the mechanical properties of the rock units that constitute the volcano. Second, the models show that dike-induced stresses and surface deformation depend much on the mechanical properties of the layers between the dike tip and the surface. In particular, the models indicate that soft layers and weak contacts between layers may suppress the dike-induced tensile stresses and the associated surface deformation. Thus, many dikes may become injected and arrested with little or no surface deformation. Generally, the numerical models suggest that standard analytical surface-deformation models such as point sources (nuclei of strain) for magma-chamber pressure changes and dislocations for dikes should be used with great caution. These models normally assume the volcanoes and rift zones to behave as homogeneous, isotropic half spaces or semi-infinite plates. When applied to stratovolcanoes composed of layers of contrasting mechanical properties and, particularly at shallow depths, weak or open contacts, inversions using these analytical models may yield results that, at best, are unreliable.
V34A-05 16:30h
The CALIPSO Borehole Project at Soufriere Hills Volcano, Montserrat, BWI: Status and Scientific Overview of Prodigious Dome Collapse of July 2003
The CALIPSO Project (Caribbean Andesite Lava Island-volcano Precision Seismo-geodetic Observatory) has greatly enhanced the monitoring and scientific infrastructure at the Soufriere Hills Volcano, Montserrat with the recent installation of an integrated array of borehole and surface geophysical instrumentation at four sites (Mattioli et al., 2004). The sensor package at each site includes: a single-component, very broad band, Sacks-Evertson strainmeter, a three-component seismometer (~Hz to 1 kHz), a Pinnacle Technologies series 5000 tiltmeter, and a surface Ashtech u-Z CGPS station with choke ring antenna, SCIGN mount and radome. The project has been successfully launched with its capture of the tremendous SHV lava dome collapse of 12-13 July 2003 (Herd et al., 2003), involving about 120 million cubic meters--the largest lava dome collapse in the historical record. A wide variety of unique geophysical signals were acquired CALIPSO instrumentation during the July 2003 collapse and important constraints on a variety of volcanic processes are being obtained. For example, tsunami waves were generated 2 km east of the volcanic dome by pyroclastic flows entering the sea. We reconstruct collapse volume-time history from seismic signals generated by pyroclastic flows, using the method of Brodscholl et al. (2000). The tsunami left flotsam strandlines of runup >8m high along the east coast of Montserrat, and waves ~0.5m high were reported from Guadaloupe. Unique borehole dilatometer data (Voight et al., 2003; Mattioli et al., 2003; 2004) record details of tsunami wave passage. One station is located 40m from the sea, with the instrument ~180m below MSL. Strain wave packets at periods of ~200-500s occurred, consistent in period and amplitude with water loading from passing tsunami waves. Wave packets between ~0600-1130 LT can be correlated with pyroclastic flow seismic data. Non-linear Boussinesq hydrodynamic modeling fits wave decay from source to instrument site and is consistent with wave periods and delay times. Coherent near-field waves depend on flow volume, submerged time of motion, and bathymetry. The model matches the delay time between pyroclastic flow entry time and arrival of tsunami waves at the instrument site.