V11C-0740
Estimated pressure source and vertical deformation in Tatun volcano group, Taiwan, detected by precise leveling in June 2006-August 2007
Tatun volcano group including more than 20 volcanoes is located in the 15 km northeastward from Taipei, Taiwan. Although Chihsing-shan, the highest peak of the Tatun volcano group, has no record of volcano eruption in history, it has a hydrothermal activity characterized by some strenuous fumarolic activities and hot springs. Seismological network installed in 2003 detects a micro-seismic activity such as the volcano-tectonic earthquakes, tremors, monochromatic events and long-period earthquakes in and around Chihsing-shan (Lin et al., 2005; Konstantinou et al., 2007). Since those volcano-seismic swarm occur just around some fumaroles, it strongly suggests that the micro- seismic activity and the hydrothermal activity are closely related. It is generally accepted that the swarm activity around volcano is often accompanied by vertical deformation (Hill et al.,2003; Oris et al.,1999; Kimata et al., 2004). The relationship between the micro-seismic activity and hydrothermal activity is an important consideration and can be investigated by using precise leveling data. In this study, therefore, we established 10km leveling route from south part to north part of the Chihsing-shan volcano to detect the vertical deformation and the leveling surveys were conducted in June 2006, March 2007 and August 2007. We detected the deformation with the maximum subsidence of 5 mm in the east part of the Chihsing-shan volcano for 9 months from June 2006 to March 2007. The result of the observation between March-August 2007 is similar to the preceding observation and the total subsidence of 10mm was detected for 14 months from June 2006 to August 2007. It suggested same mechanism is continued throughout the year and it also shows the high accuracy of our observation. Since seismic activity was low in the period of our precise leveling, we believe that volcanic fluid was not supplied to the shallow portion around the volcano in this period. The distribution of the leveling route is not good enough to estimate the exact location of pressure source and its shape based on these data and we need constrain search area using other data for getting an appropriate result. One of the leveling results, conducted for a research of Jinshan fault at the period from 2004 to 2005 by Central Geological Survey in Tiwan, shows the subsidence around Len-Shuei-Keng. We assume the spherical source might exist around Len-Shuei-Keng located at the center of the swarm area. We estimate the location of pressure source on that condition by employing a genetic algorithm (GA). In order to give better constraint to our estimation and compare the pressure-source with the swarm in detail, we made a new leveling route over a distance of 10 km on August 2007. We will be able to discuss the detail relationship between earthquake swarm and deformation in the near future.
V11C-0741
Anomalous Earthquakes Associated With Nyiragongo Volcano (D.R. Congo)
The Virunga Volcanic Complex, located in the Western Rift Valley of the East African Rift, is the site of a series of unusual earthquakes. Five anomalous events, each of moderate size (4.6<Mw<5.3), were located using long-period surface waves recorded on the global seismic network (Ekström, 2006). Despite their size, these earthquakes do not appear in standard global seismicity catalogs. Several common techniques were used to investigate the character of these seismic sources, and to determine why earthquakes of such magnitude evaded detection in an area in which many smaller events are routinely detected and located. The frequency content of these anomalous earthquakes was compared with those from local events found in global catalogs, and it was shown that these five events are greatly depleted in frequencies above 0.1 Hz. At 1 Hz, the difference in spectral amplitude between detected and undetected earthquakes of similar size and location is between two and three orders of magnitude. Centroid-moment-tensor solutions were computed for each event. The earthquakes are highly non-double-couple in nature, each having a large CLVD component of the moment tensor. Their focal mechanisms are stable and robust, and are highly irregular in character. Surface waves are also well fit by a source model composed of a series of vertical forces. The first three earthquakes occurred within days of the eruption of Nyiragongo in January 2002, while the last two events are not linked to major eruptive activity at either Nyiragongo or its neighbor, Nyamuragira. Systematic differences in frequency spectra were seen between these two groups although all five have similar focal mechanisms. These variations are potentially due to a difference in physical mechanism of the source. A variety of potential source processes are explored, including slip on a conical ring-fault, massive stoping within a buried magma chamber, and degassing/phreatomagmatic explosions.
V11C-0742
Tectonically-induced dike dilation as a model for long-duration episodes of precursory volcanic seismicity
Some explosive eruptions are preceded by episodes of volcano-seismic unrest lasting several months to a year or more. Volcanotectonic (VT) seismicity occurring during such episodes is often recorded at shallow depths many months before the eruption onset, and reflects a locally reoriented stress field suggestive of magma pressurization at shallow levels. However, the explosive nature of the ensuing eruption and other petrologic evidence often argue against prolonged shallow magma storage prior to these eruptions. Thus a new model for the generation of early precursory VT seismicity is required that accounts for all seismological and petrologic observations. It is hypothesized that a ‘passive dike opening' model may account for long-duration precursory seismicity characterized by shallow and/or random VT hypocenters and reflecting a localized ~90° stress field reorientation, without requiring ascent of magma to seismogenic levels until a later stage of precursory unrest. Using a combination of gelatin analog and numerical models, it is shown that a localized increase in the magnitude of regional maximum compressive stress in a block of crust containing a mechanically weakened zone results in dilation of a dike-like conduit through which magma can subsequently rise. Crack dilation is accompanied by a local stress field perturbation that may induce early-stage VT seismicity throughout the brittle crust but focused at shallow levels. Model results are compared to patterns of precursory VT seismicity from the 1992 eruptions of Crater Peak, Alaska, and are found to be in good agreement with observations.
V11C-0743
Characterizing Magmatic Activity at Mount Baker, Washington With Inversion of Slope Distance Data.
Surface deformation studies at active volcanoes are used to detect changes to magmatic source regions beneath the volcano. At Mount Baker, Washington, continued elevated gas (CO2 and H2S) and heat flux from fumaroles in Sherman Crater indicate the presence of a degassing magma reservoir. We assess if surface deformation has occurred on Mount Baker during the last quarter century by collecting a modern geodetic data set to compare with previous slope distance measurements acquired in 1981 and 1983 with EDM. Campaign GPS surveys in 2006 and 2007 provide slope distance measurements of all 19 trilateration lines on Mount Baker. These surveys determined that slope distances have predominantly shortened around the edifice at rates of less than 2 mm/yr. The greatest slope length change detected (HDLY-RSVT) is -17 ± 4 ppm on the northern flank of the volcano. We fit a strain model to the weighted slope change data using a nonlinear least-squares regression to characterize a two dimensional surface strain tensor. These results indicate contractional strain centered near the crater with and aerial dilation rate of less than 0.5 microstrain/yr. We also use these data to invert source parameters for a spherical magma source at depth to provide estimates of net volume and mass change of the magma reservoir. The inversion results are analyzed in conjunction with microgravity and gas flux data to better understand the current magmatic quiescence at Mount Baker.
V11C-0744
Simulation of LP Volcanic Signals Using a Fluid-Filled 3D Cylindrical System
We present the preliminary results of a dynamic model developed to study the elastic deformation of a fluid-filled 3D cavity system with finite cylindrical geometry driven by pressure gradient and changes in the velocity flow of a Newtonian fluid. We analyze the dynamic interaction of the fluid-cavity system buried in an infinite, elastic and homogeneous 3D space with physical properties similar to those encountered in volcanic environments. The basic idea is to model the opening and closure of volcanic conduits controlled by magmatic fluids ascending to the surface. The fluid-cavity system is treated within a 3D elastic space, applying the elastodynamics theory and expressing the fluid motion using the continuity equation and the conservation of momentum. The fluid and the solid are dynamically coupled in the boundary of the cavity's walls by enforcing continuity in radial stresses and velocities. The system is governed by two ordinary second order non-linear differential equations solved numerically applying the fifth order Runge-Kutta method. The solution gives the fluid's&pvertical velocity and the radial deformation of the cavity in any point and at any time. The system is triggered by a pressure excitation function initially located at the bottom of the conduit. The model allows for coupling any serial number of fluid-filled cylinders of different sizes (resembling the variable section of realistic volcanic conduits) in order to simulate long period seismic signals. Preliminary results are calculated with different systems of five coupled cylindrical cavities with different radios and lengths. Each system considers andesitic and basaltic fluid flows with properties in the range of mean values of the respective magma types. The simulated signals display characteristic waveforms and frequency content of the observed long period seismic signals in active volcanoes.
V11C-0745
Observations of Cyclic Strombolian Eruptive Behavior at Fuego Volcano, Guatemala Reflected in the Seismo-Acoustic Record
Fuego volcano is a high (3770m) basaltic cone which has been marked by >60 subplinian explosive eruptions and several year-long periods of nearly constant low level strombolian eruptions and lava flows. We have studied the current strombolian activity in detail to understand how a high cone erupts gas-rich basalt nearly continuously for many years. We compiled two years of observations of volcanic activity at Fuego in conjunction with digital seismo-acoustic recordings during a six month subset between January and July 2007. From visual observations we divide activity into three primary behaviors often occurring in a repeatable order starting with: 1) periods of lava effusion and subordinate strombolian explosions, followed by 2) paroxysmal, extended-duration, continuous strombolian emissions, and finally 3) periods of discrete degassing explosions with no lava effusion. The characteristic cycle of lava effusion, strombolian eruption, and degassing explosions repeated at least four times during the two-year study interval. Time durations of lava effusion and discrete explosive activity varied from weeks to months during 2005-2007, whereas the duration of the sustained strombolian eruptions were more consistently 24 to 48 hours. Our new seismic and acoustic data provide continuous records of eruptive behavior, which allow us to better quantify temporal duration and elucidate source processes associated with the various phases of the eruptive cycles. The seismo-acoustic station was located 7 km southwest of the active summit of Fuego, and included two infrasonic microphones and a 1 Hz short-period seismometer. These data are analyzed in conjunction with visual observations to identify spectral and amplitude characteristics of the three behaviors. For instance, we record relatively low amplitude acoustic and seismic signals and seismic tremor associated with lava effusion. We identify sustained strombolian eruptions with modulated tremor in both the seismic and acoustic wave fields. And we associate discrete degassing explosions possessing high amplitude acoustic waveforms and impulsive onsets with low amplitude seismic waveforms dominated by the ground-coupled airwave.
V11C-0746
Nature of low frequency earthquakes observed at Asama volcano, Japan: Time variation of wave parameters and hypocenter distribution
Long period (LP) events called N-type earthquakes are typical phenomena observed at many active volcanoes, such as Kusatsu-Shirane, Asama, Tokachi-dake volcanoes. They are probably related to activities of magma, ground water or volcanic gas and many source mechanisms such as resonance of fluid cracks or spheres are proposed. In this study, we analyze the LP events observed at Asama volcano in Dec. 1-10, 1996, to reveal their source process with the high quality data obtained by the seismic network close to the summit crater of the volcano. We observed 112 N-type earthquakes during the period. The waveforms of these events seem to be a quasi-monochromatic oscillation with gradually decreasing amplitude. The spectrum has a dominant peak at 1.6-7.2 Hz, most of which make a group (Group 0) in which the dominant peak changes from 2.0 Hz to 1.6 Hz gradually, indicating that the scale or the physical properties of the LP source changes gradually if we accept the resonance model. Other groups appear in Dec. 3−6 (Group 1) and in Dec. 9−10 (Group 2) in which the dominant peak changes from 7.2 Hz to 1.4 Hz and 4.3 Hz to 1.6 Hz gradually, respectively. This indicates that two or more sources of the N-type earthquakes exist simultaneously. Attenuation factors have a positive but weak correlation with the frequency of dominant peaks. Hypocenters of the events determined by the travel time of the first motion are concentrated within a depth of 300 m underneath the summit crater and are distributed in the shallow part of the region where B-type earthquakes occur. The events of Group 0 are concentrated underneath the southwest side of the crater, and the events of Group 1 and 2 distribute in the east side of the crater.
V11C-0747
A Source Mechanism of Singular Long-Period Earthquakes Observed Before the 2004 Eruptions at Asama Volcano, Central Japan
At 11:02 (GMT) on September 1, 2004, a moderate-sized summit eruption (VEI~2) occurred for the first time in the last 21 years. Before the eruption, long-period seismic signals with singular waveforms had been frequently observed at two summit stations (KAC2 and KAH2) at least since September 5, 2002. They have impulsive ground velocities with a dominant period up to 10 sec and repetitive outgoing initial motions from the vent. The signals were so feeble that we could hardly detect them even at the next closest station MAE, about 1.4km away from the summit, and thus, it was difficult to determine the location of their sources based on the conventional hypocenter determination method using travel times. The particle motions suggest that the sources are located to be relatively shallow, just beneath the bottom of the summit crater. In this study we tried to determine their source mechanism by waveform inversion. We prepared the Green's functions by the finite-difference method including 3-D topography of 10m-mesh DEM data, and used long-period band (0.05-0.25Hz) of velocity data in order to avoid error from the lack of structural information. In the inversion analyses, we can use waveform data at only 3 stations (KAC2, KAH2, and MAE): so the resolution of inversion result is poor in estimating 6 components of moment-tensor freely as superposition of short-time basis functions. Then, we made various kinds of inversions to find a possible range of solution: 6 components moment-tensor inversion with free source time function, 6 components moment-tensor inversion with fixed source time function, and fixed mechanism( sphere, cylinder, and crack) with free source time function. Source location and directions for cylinder and crack are derived by grid-search, while the other parameters are estimated by singular-value decomposition and sharp cut-off approach. The source location is estimated to be a bit southwest and about 0-100m below the bottom of summit crater. Source mechanisms derived from our various inversions almost commonly show EW-directed dipole to be prominent, suggesting this dipole component is an essential element in the source process of the singular earthquakes. Now we are trying to make a source model to explain the inversion result.
V11C-0748
The Role Played by Rainfall on the Seismicity at Fogo Volcano, Azores
Fogo volcano is an active central volcano, with a lake filled caldera, in the central part of Sao Miguel Island, Azores, whose current activity is limited to fumarolic and hydrothermal activity. It is affected by important active tectonic structures, with extremely high seismic activity and micro-seismicity concentrated in very frequent swarms. A recurrent feature of the seismicity observed in volcanic regions is the occurrence of seismic families, whose origins can be attributed to the same source mechanism, acting in the same small rock volume. Doublets/multiplets were identified in this study within a catalogue of small magnitude (usually <3) volcano- tectonic events recorded in 2003-2004 by a selection of stations around Fogo volcano. All events have been cross-correlated and pairs whose waveforms exhibited a cross-correlation coefficient equal to or higher than 0.9 were analyzed through a cross-correlation based technique (CWI, Snieder et al. 2002). Subtle velocity variations found between events highlight a seasonal cycle of the velocity patterns, with lower velocity in winter time and higher velocity during summer months. Those results, together with quantitative differences between the same doublets at different stations, exhibit an excellent correlation with the rainfall. A seasonal effect can also be broadly seen in the seismicity occurrence, and some of the swarms recorded over the two year period occur in close temporal association with higher than average rainfall. Moreover, temporal and spatial analysis of several swarms highlighted the lack of any main-shock after-shock sequence and organized migration of the hypocenters. This is suggestive of a very heterogeneous stress field. Vp/Vs is found to be lower than usually observed in volcanic areas, an occurrence likely related to the presence of fluid associated with the geothermal system (Saccorotti et al., 2004). Taken together, these two observations suggest that fluid migration/pressurization play a major role in triggering the recorded seismicity. The geothermal fluids around Fogo massif have been identified as derived from meteoric water, which infiltrates through Fogo Lake and the volcano flank and flows from south to north on the northern flank (Carvalho, 1999). These results point to the important role played by rainfall in triggering seismicity at Sao Miguel, possibly through pressure changes at depth in response to surface rain and/or interaction with the geothermal system.
V11C-0749
A Preliminary Study of Seismicity at Ceboruco, Volcano, Nayarit, Mexico
Ceboruco Volcano is located northwestern of Tepic-Zacoalco graben (Jalisco, Mexico). Its volcanic activity can be divided in four eruptive cycles differentiated by their volcano explosivity index (VEI) and chemical variations as well. As a result of andesitic effusive activity, during the first cycle the "paleo-Ceboruco" edifice was constructed. The end of this cycle is defined by a plinian eruption (VEI is estimated between 3 and 4) which occurred some 1020 years ago and formed the external caldera. During the second cycle an andesitic dome extruded in the interior of the caldera. The dome, called Dos Equis, collapsed and formed the internal caldera. The third cycle is represented by andesitic lava flows which partially cover the northern and south-southwestern part of the edifice. The last cycle is represented by historic andesitic lava flows located in the southwestern flank of the volcano. In February 2003 as part of an agreement with Nayarit Civil Defense a seismic station was installed in the SW flank of the volcano. The station is equipped with a Marslite (lennartz) digitizer with a 3DLe 1Hz. seismic sensor. Detection system is based on a STA/LTA recording algorithm. More than 2000 small earthquakes have been attributed to various local sources, and some of this earthquakes are possibly located beneath Ceboruco volcano. A preliminary classification separates high frequency and low frequency seismic events. The sources of high frequency earthquakes appear to be distributed as evidenced from waveforms variety and changing S-P arrivals separations. The low frequency seismic events also show varying signatures and some of them exhibit extended coda, including some monochromatic character.
V11C-0750
Tenerife Island Ddeformation Analysis in the 2003-2006 Period Employing the Coherent Pixel Technique and ENVISAT SAR Images
Geodetic measurements in volcanic areas are crucial as they serve as input information to the deformation analysis techniques in order to achieve pre-eruption ground displacement predictions. We present ground motion results obtained by applying the Coherent Pixel Technique, an Advanced Differential SAR Interferometry algorithm, which reveal surface deformation episodes in the Tenerife Island. The study is carried out by employing DInSAR interferograms concerning the 2003-2006 period corresponding to both ascending and descending images acquired by the ENVISAT ASAR sensor. The obtained results highlight a clear deformation areas in several parts of the island Tenerife, which is consistent with previous results.
V11C-0751
The Radon monitoring network at Stromboli Volcano: tracking precursors and earthquake- volcano interactions
We investigated radon soil degassing at Stromboli volcano since May 2002. We used E-PERM electretes (1-4 days exposure) and Track-Etch detectors (20-40 days exposure) to evaluate background levels and anomalous values in radon emissions for the operating radon network consisting of 25 stations. These values were used to construct keys maps that identify radon anomalies due to changes in volcanic activity and/or the occurrence of major seismic events. Higher emissions are concentrated at the summit and along the NE sector of the island (controlled by faults trending N40 and N60). Automated measurements started May 2005 at two distinct summit sites. Data processing outline a general correlation between atmospheric pressure and radon degassing. Moreover, comparison of our data with those collected by INGV on carbon dioxide shows a very strong correlation, thus further supporting the idea that CO2 is an active "radon carrier". Measurements of the radon progeny (218Po and 214Po) give us a clue to discriminate between the residing gas in the porous medium and the "new coming" 222Rn parent. Under isotopic equilibrium conditions the ratio 222Rn/214Po should be equal to unity. Conversely under dynamic conditions, an increase in radon fluxing (from below) will result in an increase of 222Rn with respect to 214Po. Activity versus time graphs show that major explosions are preceded by the crossing over of 222Rn onto the 214Po curve. We finally analyzed the link between radon emissions, volcanic tremor and number of VLP events. The overall data show a general time-correlation among these parameters, and further indicate that radon is a key parameter in volcano suirvellance
V11C-0752
Moment Tensor Inversion of Explosive Events at Arenal Volcano, Costa Rica.
One of the fundamental aims of volcano seismology is to characterise the magmatic system and to determine the evolution of this system. The analysis of seismic waves is used to determine the internal structure of volcanoes and to determine the source processes. By understanding the source processes we can better constrain the evolution of the volcano plumbing system. Arenal volcano is a small basaltic andesite stratovolcano in north- western Costa Rica. An experiment including the deployment of ten Guralp CMG40T seismometers around the volcano was carried out in 2005. This temporal network recorded several explosive events. A waveform inversion is performed to retrieve the moment tensor associated with the source of these explosions. The Green's functions were calculated using an elastic lattice wave simulation method including a layered velocity model constrained by the latest published velocity models and Arenal topography. We will present preliminary results of our source inversion and our interpretation of these results.
V11C-0753
Autonomous Triggering of in situ Sensors on Kilauea Volcano, HI, from Eruption Detection by the EO-1 Spacecraft: Design and Operational Scenario.
Response time in acquiring sensor data in volcanic emergencies can be greatly improved through use of autonomous systems. For instance, ground-based observations and data processing applications of the JPL Volcano Sensor Web have promptly triggered spacecraft observations [e.g., 1]. The reverse command and information flow path can also be useful, using autonomous analysis of spacecraft data to trigger in situ sensors. In this demonstration project, SO2 sensors have been incorporated into expendable "Volcano Monitor" capsules to be placed downwind of the Pu'U 'O'o vent of Kilauea volcano, Hawai'i. In nominal (low) power conservation mode, data from these sensors are collected and transmitted every hour to the Volcano Sensor Web through the Iridium Satellite Network. If SO2 readings exceed a predetermined threshold, the modem within the Volcano Monitor sends an alert to the Sensor Web, triggering a request for prompt Earth Observing-1 ( EO-1) spacecraft data acquisition. During pre-defined "critical events" as perceived by multiple sensors (which could include both in situ and spaceborne devices), however, the Sensor Web can order the SO2 sensors within the Volcano Monitor to increase their sampling frequency to once per minute (high power "burst mode"). Autonomous control of the sensors' sampling frequency enables the Sensor Web to monitor and respond to rapidly evolving conditions before and during an eruption, and allows near real-time compilation and dissemination of these data to the scientific community. Reference: [1] Davies et al., (2006) Eos, 87, (1), 1&5. This work was performed at the Jet Propulsion Laboratory-California Institute of Technology, under contract to NASA. Support was provided by the NASA AIST program, the Idaho Space Grant Consortium, and the New Mexico Space Grant Program. We thank the personnel of the USGS Hawaiian Volcano Observatory for their invaluable assistance.
V11C-0754
Volcanic Surface Deformation in Dominica From GPS Geodesy: Results From the 2007 NSF- REU Site
GPS measurements have been collected on the island of Dominica in the Lesser Antilles between 2001 and 2007, with five month-long campaigns completed in June of each year supported in part by a NSF REU Site award for the past two years. All GPS data were collected using dual-frequency, code-phase receivers and geodetic-quality antenna, primarily choke rings. Three consecutive 24 hr observation days were normally obtained for each site. Precise station positions were estimated with GIPSY-OASISII using an absolute point positioning strategy and final, precise orbits, clocks, earth orientation parameters, and x-files. All position estimates were updated to ITRF05 and a revised Caribbean Euler pole was used to place our observations in a CAR-fixed frame. Time series were created to determine the velocity of each station. Forward and inverse elastic half-space models with planar (i.e. dike) and Mogi (i.e. point) sources were investigated. Inverse modeling was completed using a downhill simplex method of function minimization. Selected site velocities were used to create appropriate models for specific regions of Dominica, which correspond to known centers of pre-historic volcanic or recent shallow, seismic activity. Because of the current distribution of GPS sites with robust velocity estimates, we limit our models to possible magmatic activity in the northern, proximal to the volcanic centers of Morne Diablotins and Morne aux Diables, and southern, proximal to volcanic centers of Soufriere and Morne Plat Pays, regions of the island. Surface deformation data from the northernmost sites may be fit with the development of a several km-long dike trending approximately northeast- southwest. Activity in the southern volcanic centers is best modeled by an expanding point source at approximately 1 km depth.
V11C-0755
Inflation of Kilauea Along the Southwest Rift Zone in 2006
We report on InSAR and GPS results showing the first crustal inflation along the southwest rift zone at Kilauea volcano in over 30 years. Two independent interferograms (May 2 to August 2, 2006 and June 22 to Nov 7, 2006) from the ALOS PALSAR instrument reveal domal uplift located southwest of the main caldera. The uplift is bounded on the northeast by the caldera and follows the southwest rift zone for about 12 km. It is approximately 8 km wide. We use data derived from permanent GPS stations to calibrate the InSAR displacement data and estimate uplift of 8.3 cm during the first interferogram and 8.7 cm during the second with volumes of 2.8x106 m3 and 2.4x106 m3 respectively. The earthquake record for the periods before, during, and after inflation shows that a swarm of shallow earthquakes (z < 5 km) signaled the beginning of the uplift and that elevated levels of shallow seismicity along the rift zones occurred throughout the uplift period. GPS data indicate that the inflation occurred steadily over nine months between mid-January and mid-October, 2006 making injection of a sill unlikely.
V11C-0756
The Influence of Near Surface Velocity Structures on Long-Period (LP) Volcano-Seismic Signals: Examples Using Simulations from Mt Etna.
Long Period (LP) signals with dominant periods in the range 0.2-2 Hz have received particular attention on volcanoes, as they are thought to be directly associated with moving fluids or resonating fluid-filled conduits. Consequently, in an effort to better understand fluid-driven processes, inverting for LP source mechanisms is becoming increasingly common. As the majority of LP events occur at shallow depths & have short path lengths and km-long wavelengths, the role of edifice heterogeneity is usually ignored. We use 3D full wavefield simulations in heterogeneous Mt Etna models with topography to generate a wavefield radiated from a synthetic LP source. Greens functions are calculated for a homogeneous model with topography. Moment Tensor plus single force inversions of these synthetics demonstrate the extreme sensitivity of the solution for LP source forces to near-surface volcano structure. In particular, spurious forces and incorrect source orientation are obtained if the top 400 m is poorly constrained. Sensitivity kernels help to elucidate the details of the propagation effects which lead to poor source characterization and allow us to suggest an improved protocol for the inversion of real LP signals. We conclude that the effect of poorly resolved velocity structure can be minimized by using a priori information about the source geometry and inverting signals for as few free model parameters as possible.
V11C-0757
Preliminary Results From a Temporary Seismic Network at Mt. Baker, Washington
In August 2007, a temporary network of four broadband seismometers was deployed at Mt. Baker, WA. This marks the first time that a seismic network has been located on Mt. Baker since a period of volcanic unrest in 1975-76. These data can therefore provide a new window into Baker's background seismicity. The network was composed of four Guralp CMG-6TD seismometers, each of which was deployed for 4-8 weeks between August and October 2007. Instruments were located between 0.5 and 10 km from the volcano's summit. Up to 10 earthquakes per day were recorded across the network, although the majority of these events are thought to be associated with ice rather than volcanic processes. Volcanic earthquakes appear to be infrequent, consistent with previous interpretations of Baker seismicity; however, the short duration of the deployment makes it difficult to extrapolate to longer time scales. Other signals recorded during the deployment include rockfalls from the rim of Sherman Crater, regional and teleseismic earthquakes, and a deep long-period earthquake located to the SE of the volcano. Because there is only a single permanent seismometer located within 30 km of Mt. Baker, hypocenters for previous Baker earthquakes have been poorly constrained for all but the largest (>M2) events. This network provides the first opportunity in 30 years to study microseismicity at Mt. Baker and examine its background behavior.
V11C-0758
The Magmatic Component of the Plate Boundary Observatory.
The Plate Boundary Observatory (PBO) component of the NSF-funded Earthscope program has a significant complement of instruments devoted to the study of magmatic systems. There are ten target areas: Akutan, Unimak, Augustine, Mt St Helens, Long Valley, Yellowstone, Lake Tahoe, Medicine Lake, Mt Lassen, and Mt Shasta that include 22 borehole strainmeters, 22 borehole seismometers, 26 borehole tiltmeters and 110 continuous GPS stations all returning data in near real-time. In conjunction with the existing instrumentation operated by the USGS Volcanic Hazards Programs Volcano Observatories, this represents a significant array of tools for exploring various volcanic processes. In the first four years of the project, PBO has captured two volcanic events (Mt St Helens and Augustine) far exceeding the anticipation of capturing one event in the first 15 years of the project. This presentation gives an overview and status of the program, the various targets, instrumentation and results. http://pboweb.unavco.org/
V11C-0759
Edifice Instability and Basal Fault Friction
Edifice instability, which is a common feature of volcanoes, strongly affects eruptive phenomena and evolution of the volcano edifice. We investigate the interplay of several factors that determine sector failure of axi-symmetric volcanoes including slope of the volcano, the dip of the basal fault, sector angle, and effect of magmatic intrusion into flank rift zones. We employ a simple stability model of volcano sector failure in which failure is resisted by friction on the basal fault. In general, we find that axi-symmetric volcanoes are less stable, or require a higher coefficient of basal fault friction to resist failure, than Hawaiian-type linear rift volcanoes. Sector failures commonly define angles in the range 90°-120° - this is consistent with the model results, which show that larger sector angles to 180° are less favorable for failure. Also, negatively dipping basal faults are more stable than a horizontal or positively dipping fault. We apply this model to estimate the effective coefficient of friction of volcanoes built on a weak substrate such as Concepcion and Maderas in Nicaragua.
V11C-0760
Low-Cost Photogrammetric Technique Used to Measure Dome Growth at Mount St. Helens Volcano, 2007-2007
We describe a low-cost application of digital photogrammetry using commercial grade software, an off-the-shelf digital camera, a laptop computer and oblique photographs to reconstruct volcanic dome morphology during the on-going eruption at Mount St. Helens, Washington. Renewed activity at Mount St. Helens provides a rare opportunity to devise and test new methods for better understanding and predicting volcanic events, because the new method can be validated against other observations on this well-instrumented volcano. Uncalibrated, oblique aerial photographs (snap shots) taken from a helicopter are the raw data. Twelve sets of overlapping digital images of the dome taken during 2004-2007 were used to produce digital elevation models (DEMs) from which dome height, eruption volume and extrusion rate can be derived. Analyses of the digital images were carried out using PhotoModeler software, which produces three dimensional coordinates of points identified in multiple photos. The steps involved include: (1) calibrating the digital camera using this software package, (2) establishing control points derived from existing DEMs, (3) identifying tie points located in each photo of any given model date, and (4) identifying points in pairs of photos to build a three dimensional model of the evolving dome at each photo date. Text files of three-dimensional points encompassing the dome at each date were imported into ArcGIS and three-dimensional models (triangulated irregular network or TINs) were generated. TINs were then converted to 2 m raster DEMs. The evolving morphology of the growing dome was modeled by comparison of successive DEMs. The volume of extruded lava visible in each DEM was calculated using the 1986 pre-eruption crater floor topography as a basal surface. Results were validated by comparing volume measurements derived from traditional aerophotogrammetric surveys run by the USGS Cascades Volcano Observatory. Our new "quick and cheap" technique yields estimates of eruptive volume consistently within 5% of the volumes estimated with traditional surveys. The end result of this project is a new technique that provides an inexpensive, rapid assessment tool for tracking lava dome growth or other topographic changes at restless volcanoes.