V11B-1419 0800h
Small inflation source producing seismic and infrasonic pulses without eruptions, during the 2000 Miyake-jima volcanic activity, Japan
We detected low frequency seismic events and infrasonic pulses that repeated during the 2000 activity of Miyake-jima volcano, Japan. The signals are interpreted to have been excited by underground small inflation beneath a summit crater without remarkable eruptions. The low frequency seismic signals were intermittently observed with gradual increase of the magnitude toward the explosive eruption on 18 August. Small infrasonic pulses (IS pulses) occurred associated with the seismic events without eruptions, and had the pulse width of 1 to 2 seconds. The sources of IS pulses were located in the caldera area. The initial motions of low frequency events are of compression at all seismic stations, and their particle motions are outward from the region beneath the summit crater. Grid search by waveform inversion method shows that the initial motions are excited by isotropic inflation source beneath the south edge of the caldera at 1.4 km in depth. There were clear differences of the origin times between seismic events and IS pulses which emitted on the surface: 3.5 to 4.5 sec. The time difference suggests that a small isotropic inflation at the depth of 1.4 km excited a seismic pulses, simultaneously a wave that travels as a sound wave in the conduit and produce an IS pulses on the surface. In this scheme, the conduit should have been filled with a mixture gas composed of vapor and something, e.g. SO2 gas or basaltic ash, in order to explain the time difference. Our analyzing low frequency seismic signals have similar features often observed in volcanic explosions at various volcanoes; they are of compression in initial motions, and the sources are at the depth of a few km beneath the crater bottom. Repetitive events have high similarity. and infrasonic pulses were observed just after the seismic signal. However, the distinctive difference is that the seismic/airborne events did occur without remarkable eruptions. We think that we detected small underground eruptions not leading to volcanic explosions on ground surface, so to speak, seeds of eruptions.
V11B-1420 0800h
Strombolian thermodynamics from FLIR imagery: New insights on explosive styles and source conditions
We use temperature-calibrated thermal video to constrain the physical characteristics and dynamics of Strombolian eruptions. To do this, we utilize data from a Forward Looking Infrared Radiometer (FLIR) camera, operating at up to 30 Hz. 344 eruptive events at Stromboli volcano, Italy, were imaged during four field seasons spanning four years. We record a wide range of eruptive styles at Stromboli, something hitherto undocumented in the literature because of the small data sets used by previous imaging studies. Eruptions at Stromboli can generally be classified into two groups: Type 1 eruptions, which consist of spatter and an ash-poor gas plume, and Type 2 eruptions, which consist of an optically-thick, ash-rich plume, with (Type 2a) or without (Type 2b) additional spatter. This ash content relates to changes in the thermodynamics of the eruption, such as velocity, pressure, and thermal history. Because of this influence, the time-varying eruption style has ramifications for interpreting data from monitoring systems, including infrared spot-radiometry and infrasound. As a given eruption style (Type 1 vs. Type 2) is often sustained for days to weeks, the variations in these eruptive parameters, on these time scales, can relate to changes in eruptive behavior.
V11B-1421 0800h
Experiments on Porous Flow With Cooling Boundary
Fluid migrations in geological problems (hydrothermal systems, magma migration, etc.) are not purely mechanical problems. They involve heat transfer and/or chemical reactions, which may cause the structural change of fluid paths. When the fluid flows into cooler regions, it may precipitate dissolved materials along its paths. Such a structural change affects the fluid flow and leads to a temporal change in flux. In order to understand the fluid migration with thermal and chemical processes, we have conducted experiments of porous flow with a cooling boundary. We employ packed nylon beads (D=0.4 mm, L=0.4 mm) and NH$_{4}$Cl aqueous solution as a porous matrix and a fluid, respectively. The solubility of NH$_{4}$Cl is very sensitive to the temperature. A small temperature drop makes a considerable amount of precipitation. Undersaturated NH$_{4}$Cl solution is injected at a constant rate from the column of beads. The top part of the column is cooled from the outside, where NH$_{4}$Cl can precipitate in pores to reduce the permeability. We change the temperature of the cooling boundary, the fluid injection rate, and the concentration of NH$_{4}$Cl, and study the temporal change of the temperature of the fluid, the outflow rate, and the fluid pressure. We have observed three different types of temporal change. When the cooling of the fluid is not effective, the fluid does not precipitate dissolved NH$_{4}$Cl. The outflow rate and the fluid pressure are kept constant. On the other hand, when the cooling is effective, the fluid precipitates NH$_{4}$Cl. The fluid pressure increases to keep the outflow rate constant. The second case is subdivided into two types. When the cooling effect is strong, the precipitation makes a strong lid at the cooling boundary. The fluid pressure increases steadily. When the cooling effect is moderate, the elevated fluid pressure breaks the low permeability lid. The fluctuation of the fluid pressure is observed.
V11B-1422 0800h
Ground-Based Thermal Analysis of Degassing Activity at Pu'u O'o
For the last 4 years the ground-based thermal monitoring system at Pu'u O'o (Kilauea, Hawaii) has allowed a complete and continuous description of the degassing and thermal activity. The monitoring system consists on 3 Omega IR thermometers, sensitive between 8-14 microns. A 60$\deg$ aperture sensor tracks a large area (FOV $\sim$ 150 m diameter) across the crater floor, thus providing a thermal over-view of the crater floor, while two 1$\deg$ radiometers (FOV $\sim$ 2.5 m) allow detailed detection of degassing activity from single vents. Ground-based thermal observations reveal that degassing at Pu'u O'o consists mainly in sustained degassing, where gas is continuously emitted as intermittent gas puffs, and gas pistoning events. The continuous record points to systematic fluctuations of thermal data. Here during the period Jan 2001-Dec 2003 activity at Pu'u O'o has fluctuated between periods of high and low rates of thermal activity. This is reflected generally across the whole system, so that multiple vents show similar long-term fluctuations. At shorter time scales differences in the degassing mode between vents may be observed, with gas pistoning events concentrated mainly at one vent and sustained degassing at the others. Gas pistoning at Pu'u O'o is recorded thermally as an abrupt increase in temperature ($\sim$ 50-150 $\deg$C), which lasts $\sim$ 100-300 sec. The events are always preceded by a significant decrease in temperature, which points to a net drop in degassing. This is consistent with seismic data, which show a contemporary decrease in seismic energy before gas pistoning events. Statistical analysis over $\sim$ 5000 gas pistoning events extracted over the four year-monitoring reveals a typical recurrence of $\sim$ 8-9 minutes. Stacking of the events reveals two different types of thermal waveform associated with gas pistoning events: the first consists of a single thermal peak, with an amplitude of $\sim$ 50-100 $\deg$C, and lasts 100-200 sec, while the second lasts longer ($\sim$ 200-300 sec), and shows a precursory thermal peak immediately followed by a second major pulse ($\sim$ 50-150 $\deg$C). Such events are usually followed by periods of sustained degassing. The two different waveforms of gas pistoning appear to correlate with periods of low and high degassing respectively. Sustained degassing activity at Pu'u O'o is also recorded by the 1$\deg$ FOV radiometers as intermittent, high frequency, thermal pulses, which reflect intermittent gas puffing. Here our analysis focuses mainly on puffing frequency, rather than on thermal amplitude, where the latter is strongly influenced by external conditions (visibility, targeting of the sensor, etc.), while the former is a direct expression of the source process. Our investigation points to long term fluctuations in thermal activity at Pu'u O'o, where intermittent gas puffing and occurrence of gas pistoning events are the main features of degassing activity and may reflect fluctuations in the gas and mass fluxes, or a shallow system mechanism of collection and release of gas slugs. The detailed analysis of these different degassing modes, and relationships between them, is thus critical to gain new information on the degassing processes at Pu'u O'o.
V11B-1423 0800h
A Volcano Reawakens: Multiparameter Observations of Activity Transition at Galeras Volcano (Colombia)
Following 11 years of relatively low activity, Galeras Volcano (Colombia) produced a sequence of ash eruptions in July and August, 2004. Starting in March 1996, a multiparameter real-time monitoring system was installed at Galeras, as a part of a cooperative program between INGEOMINAS (Colombia) and the BGR (Germany). Broadband seismometers were installed first, with electromagnetic (EM) sensors, sensors for the chemistry and physics of the fumarole gases and a weather station following later. The data from these instruments augment the short-period seismic network and tiltmeters of Observatorio Vulcanol\'ogico de Pasto (OVP). Additional spot measurements are made using visual inspection from the ground or helicopter, a thermal camera and regular geological forays onto Galeras' slopes. Initial evidence of the activity transition appeared in the gas measurements in early June, followed by a strong increase in the shallow seismic activity below the active cone on June 27. As in many cases at other volcanoes, the most clear evidence for the transition came in the form of seismic swarms and tremor. The current activity has culminated in two brief episodes of ash emission, on July 16 and July 21, followed by two longer episodes, July 27 - August 8, and August 11 - 19. This last episode began with a large explosion and released more ash than any individual episode from 1989 to 1993. Sudden deformation, as well as changes in the electric and magnetic fields at the crater EM station, and the gas parameters such as CO$_2$ concentration and fumarole temperature accompanied the ash emissions on July 16 and July 21. Unfortunately, the EM and gas instruments were lost to ashfall shortly afterward. As of the writing of this abstract, the ash emissions continue intermittantly.
V11B-1424 0800h
Coupling Thermal and Chemical Signatures of Crustal Magma Bodies: Energy-Constrained Eruption, Recharge, Assimilation, and Fractional Crystallization (E'RA$\chi$FC)
Energy-Constrained Eruption, Recharge, Assimilation and Fractional Crystallization (E'RA$\chi$FC) tracks the evolution of an open-system magmatic system by coupling conservation equations governing energy, mass and species (isotopes and trace elements). By linking the compositional characteristics of a composite magmatic system (host magma, recharge magma, wallrock, eruptive reservoir) to its mass and energy fluxes, predictions can be made about the chemical evolution of systems characterized by distinct compositional and thermal characteristics. An interesting application of E'RA$\chi$FC involves documenting the influence distinct thermal regimes have on the chemical evolution of magmatic systems. Heat transfer between a magma-country rock system at epizonal depths can be viewed as a conjugate heat transfer problem in which the average country rock-magma boundary temperature, Tb, is governed by the relative vigor of hydrothermal convection in the country rock vs. magma convection. For cases where hydrothermal circulation is vigorous and magmatic heat is efficiently transported away from the boundary, contact aureole temperatures (~Tb) are low. In cases where magmatic heat can not be efficiently transported away from the boundary and hydrothermal cells are absent or poorly developed, Tb is relatively high. Simultaneous solution of the differential equations governing momentum and energy conservation and continuity for the coupled hydrothermal-magmatic conjugate heat transfer system enables calculation of the characteristic timescale for EC-RAFC evolution and development of hydrothermal deposits as a function of material and medium properties, sizes of systems and relative efficiency of hydrothermal vs. magmatic heat transfer. Characteristic timescales lie in the range 10$^{2}$-10$^{6}$ yr depending on system size, magma properties and permeability among other parameters. In E'RA$\chi$FC, Tb is approximated by the user-defined equilibration temperature, Teq, which is the temperature at which all parts of the composite magmatic system achieve thermal equilibrium. Comparison of the results of three EC-AFC simulations at different Teq ($1150\deg$C, $1050\deg$C, $1000\deg$C) for a mafic magma intruding middle-upper crust of mafic-intermediate composition illustrate the distinctions that can be imparted by a range of thermal regimes. Model parameters relevant to the following results include: initial Sr concentration, isotope composition and bulk D for host magma are 700 ppm, 0.7035, and 1.5, respectively; those for wallrock are 230 ppm, 0.7100, 0.05. The $1150\deg$C case (i.e., high Tb) yields the least crust-like Sr isotope signatures. The mass of wallrock that reaches thermal equilibrium is relatively small (0.26, normalized to the mass of initial host magma), although the degree of melting is high (97%). In contrast, the $1000\deg$C case (i.e., low Tb) yields the most crust-like Sr isotope signatures. This case is also characterized by the largest mass of wallrock (0.98, normalized to the mass of initial host magma) that achieves thermal equilibrium, but the degree to which this wallrock melts is small (10%). A fundamental issue that derives from these results is the relationship between the chemical evolution of the hydrothermal system and the chemical evolution of associated melt and cumulates. In particular, to what extent can predictions be made from the thermal interactions between magma and wallrock on the chemical signatures of the associated magmatic rocks and hydrothermal deposits?
V11B-1425 0800h
Comparative investigation of soil CO$_{2}$ flux measurements and geostatistical estimation methods on Masaya volcano, Nicaragua
Measurements of soil CO$_{2}$ flux (F$_{CO2}$) and its natural spatial and temporal variability in volcanic and hydrothermal environments are important for volcano monitoring, geothermal exploration, delineation of fault and fracture zones, and estimation of the contribution of CO$_{2}$ from volcanic and hydrothermal sources to the global carbon cycle. However, the choice of measurement and geostatistical methodologies may affect individual F$_{CO2}$ measurements and characterization of their natural spatial and temporal variability, the total CO$_{2}$ emission rate estimated, and the ability to assess the uncertainty associated with this estimate. We present a comparative study of F$_{CO2}$ measured by five research groups (Groups 1-5) at the 2003 IAVCEI-CCVG Eighth Workshop on Volcanic Gases on Masaya volcano, Nicaragua. Groups 1-5 measured F$_{CO2}$ using the accumulation chamber method at 5-m spacing within a 900~m$^{2}$ grid during a morning (AM) period. These measurements were repeated by Groups 1-3 during an afternoon (PM) period. Measured F$_{CO2}$ ranged from 218 to 14,719~g~m$^{-2}$d^{-1}$. The variability of the five measurements made at each grid point ranged from $\pm$5 to 167%; however, the arithmetic means of fluxes measured over the entire grid and associated total CO$_{2}$ emission rate estimates varied between groups by only $\pm$22%. All three groups that made PM measurements reported an 8-19% increase in total emissions over the AM results. Based on a comparison of measurements made during AM and PM times, this change is likely due in large part to natural temporal variability of gas flow, rather than to measurement error. To estimate the mean and associated CO$_{2}$ emission rate of one data set and to map the spatial F$_{CO2}$ distribution, we compared six geostatistical methods: arithmetic and minimum variance unbiased estimator means of uninterpolated data, and arithmetic means of data interpolated by the multiquadric radial basis function, ordinary kriging, multi-Gaussian kriging, and sequential Gaussian simulation methods. While the total CO$_{2}$ emission rates estimated using the different techniques only varied by $\pm$1.1%, the F$_{CO2}$ maps showed important differences. We suggest that the sequential Gaussian simulation method yields the most realistic representation of the spatial distribution of F$_{CO2}$, but a variety of geostatistical methods are appropriate to estimate the total CO$_{2}$ emission rate from a study area, which is a primary goal in volcano monitoring research. Acknowledgement: Part of this work was completed at Lawrence Berkeley National Laboratory, under U.S. Department of Energy Contract No. DE-AC03765F00098.
V11B-1426 0800h
Evolution of Magma Plumbing System Between Large Pyroclastic Eruption Cycles in Aso Volcano, Japan
Aso volcano is located in central Kyushu, SW Japan. Volcanic activity at Aso is characterized by four large pyroclastic eruptions, which are called as Aso-1 (occurred in 300 ka), 2 (150 ka), 3 (120 ka) and 4 (90 ka), and many small eruptions between them. Previous works have reported that the four large eruptions were basically eruptions from chemically zoned chamber of rhyolitic and basaltic magmas and their petrological features of these magmas systematically change with time; incompatible elements (e.g. $K_2$O) for a given Si$O_2$ content decrease and oxidation state increase from Aso-2 to Aso-4. By contrast, the small eruptions were mostly Plinian eruptions of silicic magmas, and their petrological features are not fully investigated. In this study, we investigate petrological features of the ejecta of the small eruptions between Aso-2 and 3 (referred to as SE-3/2) and between Aso-3 and 4 (SE-4/3) in detail and compare them with those of Aso-2, 3 and 4. The variations of SE-3/2 and SE-4/3 magmas have two common petrological features. First, $K_2$O for a given Si$O_2$ content irregularly change with time; $K_2$O for a given Si$O_2$ in SE-3/2 magmas varies in the range between that of Aso-2 and Aso-3, whereas that in SE-4/3 magmas varies in the range between that of Aso-2 and Aso-4. This means that the range of the compositional variation increases with time so that the amount of incompatible elements in the most depleted magma decreases with time. Second, oxidation state of SE-3/2 magmas is similar to that of Aso-3, and likewise, that of SE-4/3 is similar to Aso-4. The compositional variation of the magmas cannot be explained by a single fractional crystallization process, suggesting that the ejecta of the small eruptions do not come from a single magma chamber. The temporal variation of the incompatible elements may indicate that the silicic magmas are generated by partial melting of the crust due to repeated injections of basaltic magmas; the crust becomes more depleted in incompatible elements with time as a whole, whereas the degree of depletion varies spatially in the crust. The oxidation state of the whole magma system changes immediately after the large eruptions.
V11B-1427 0800h
Three-dimensional P-wave Velocity Structure of Tungurahua Volcano, Ecuador
Tungurahua Volcano in the Ecuadorian Andes is a large andesitic stratovolcano (5023 m) that has been erupting mildly since 1999. We studied the three-dimensional (3-D) P-wave velocity (Vp) structure beneath the volcano down to 5 km below sea level. We inverted 1708 P-wave first arrival times from 263 volcano-tectonic (VT) earthquakes recorded by 5 to 10 1 Hz vertical seismic stations on the volcano from August 1999 to May 2003. A non-linear inversion method was used to image the velocity structure, in which first-arrival times were calculated with a finite-difference method. The hypocenters in our model are tightly clustered and aligned roughly along a vertical structure at depths between sea level and the summit crater. A high-velocity zone exists above the central base of the volcano under the vertically aligned hypocenters, and may be interpreted as the source zone for recharge of the shallow magmatic system. High-velocity zones are also identified under the lower northeastern and southern flanks of the edifice. The northeast anomaly extends down to 1 km below sea level and possibly represents an old volcanic conduit. The southern high-velocity anomaly is underlain by a relatively low-velocity zone and its geometry suggests an old lateral dike system connected at depth to the central conduit. Except for these high-velocity zones in the central, northern, and southern flanks, the volcanic edifice is composed of low-velocity materials down to a depth of 2 km above sea level. These low velocities correlate with young unconsolidated deposits, and older highly fractured and/or altered volcanic materials.
V11B-1428 0800h
Seismic recordings of debris avalanches on Iliamna Volcano, Alaska
Seismic recordings of three debris avalanches on Iliamna Volcano show a prolonged precursory phase of activity that may represent the initial stages of failure before the onset of the avalanche. Iliamna is an andesitic stratovolcano in the Cook Inlet region of Alaska with no documented eruptions in historic times. The upper flanks of the volcano are hydrothermally altered and prone to failure. In 1994 and 1997, rock and ice debris avalanches occurred on Iliamna's eastern flank, below the main fumarole field, each with a volume of $\approx$15 x 10$^{6}$ m$^{3}$. A similar event on February 9 2004 occurred on the western side of the edifice, also near a hydrothermally altered zone. All three avalanches were detected by a short-period seismic network operated by the Alaska Volcano Observatory. Although the avalanches differ in location and volume, they produced similar seismic signals, suggesting that the events share common failure mechanics. Each signal begins with a sequence of earthquakes lasting 1-2 hours with spectral peaks between 2-8 Hz. The discrete events associated with each avalanche are similar in time series, suggesting they share a common source location and mechanism. High-resolution relocation of these events confirms that for each avalanche, the earthquakes occur at shallow depth beneath the associated avalanche headwall. In all cases the discrete events continue for up to two hours before the avalanche onset, occurring at shorter and shorter intervals, as event amplitudes decline. Eventually, individual shocks cannot be resolved and the signal degrades to diffuse continuous ground shaking. Each sequence ends with a signal comprising several broadband (2-30 Hz) pulses lasting 1-10 minutes that saturate nearby ($<$ 5 km) stations. The wide frequency band and spindle shape of the final period resembles other rockfall and avalanche signals. The diffuse and broadband periods are thus proposed to represent the avalanche, while the discrete earthquakes represent the initial stages of failure. Identification and recognition of avalanche events is critical for identifying avalanche hazards as well as discriminating between avalanche and eruption signals. The long duration of the seismic sequence suggests that some warning may be possible for future avalanche events.
V11B-1429 0800h
Tornillos and Other Volcanic Tremors in Tatun Volcanoes, Northern Taiwan
This is the first time to report several types of volcanic signals such as Tornillos (screws), harmonic signals (drops) and continuously short-period volcanic tremors at the Tatun volcanic group (Taiwan), which was usually considered as extinct because there was no any historical eruption. These volcanic signals are often reported in the active volcanoes in the earth. In particular, the tornillos have been considered as the potential precursor for volcanic eruption at Galeras volcano, Colombia. Combining all of these volcanic signals with the presviously geochemical analyses from the Helium isotop ratio, we strongly suggest that volcanic activities in the Tatun volcanic area might not be totally extinct yet and further investigations have to been done for evaluation of potential volcanic activities because the Tatun volcano group is not only just nearby two nuclear power plants but also about 15 km north to Taipei, the capital of Taiwan, in which more than seven million people live on.
V11B-1430 0800h
Seismic exploration of Fuji volcano with active sources in 2003
Fuji volcano (altitude 3,776 m) is the largest basaltic stratovolcano in Japan. In late August and early September 2003, seismic exploration was conducted around Fuji volcano by the detonation of 500 kg charges of dynamite to investigate the seismic structure of that area. Seismographs with an eigenfrequency of 2 Hz were used for observation, positioned along a WSW-ENE line passing through the summit of the mountain. A total of 469 observation points were installed at intervals of 250-500 m. The data were stored in memory on-site using data loggers. The sampling interval was 4 ms. Charges were detonated at 5 points, one at each end of the observation line and 3 along its length. The first arrival times at each observation point for each detonation were recorded as data. The P-wave velocity structure directly below the observation line was determined by forward calculation using the ray tracing method [Zelt and Smith, 1992]. The P-wave velocity structure below the volcano, assuming a layered structure, was found to be as follows. (1) The first layer extends for about 40 km around the summit and to a depth of 1-2 km. The P-wave velocity is 2.5 km/s on the upper surface of the layer and 3.5 km/s on the lower interface. (2) The second layer has P-wave velocities of 4.0 km/s on the top interface and 5.5 km/s at the lower interface. The layer is 2<ETH>5 km thick to the west of the summit and 1-2 km thick to the east, and forms a dome shape with a peak altitude of 2000 m directly below the summit. (3) The third layer is 5-12 km thick and has P-wave velocities of 5.7 km/s at the top interface and 6.5 km/s at the lower interface. This layer reaches shallower levels to the east of the summit, corresponding to the area where the second layer is thinner. Mt. Fuji is located slightly back from where the Philippine Sea Plate subducts below the Eurasian plate in association with collision with the Izu Peninsula. Matsuda (1971) suggested that Mt. Fuji lies on the same uplifted body as the Tanzawa Range to the east. This uplifted body is formed by plate subduction and collision with the Izu Peninsula, and is believed to have influence at significant depth. This is considered to be the reason for the change in the geologic structure beneath Fuji volcano from west to east. The dome structure of the bedrock layer (second layer) directly beneath the summit is considered to have formed in the initial period of volcanic activity that formed Mt. Fuji, leading to the subsequent formation of Komitake volcano, Ko-Fuji volcano and the present day Fuji volcano.
V11B-1431 0800h
Monitoring and Studies on High Risk Volcanoes in the Java Region/Indonesia
In co-operation with several Indonesian science institutions a programme was launched in early 2004 to study the various expressions of volcanism in the Java region, its relations to the active subduction environment, and to install a real-time multi-parameter (MP) monitoring station. Three Indonesian volcanoes were seleced: Krakatau, Merapi and Kelut. The interdisciplinary programme consists of three sub-projects: (1) Krakatau Monitoring (KRAKMON): real-time observations of the volcano noise and micro-seismicity, of electro-magnetic, gas-chemical and thermal parameters as well as of the deformation of the volcanic edifice and meteorological influences. Video monitoring and a data centre are integrated. The station will be set into operation in April 2005. (2) Merapi Amphibious Experiment (MERAMEX): tomographic studies (vp and Qp) on the Merapi edifice and its surroundings by active and passive seismic surveys simultaneously along with 120 land stations and 14 Ocean Bottom Sensors providing high-resolution 3-dimensional models from the surface down to the plate interface in order to image the complete pathways of fluids and melts. Field work ended in October 2004. First results are presented. (3) Development of Highly Explosive Volcanoes at Active Continental Margins (DEVACOM): Analysis of samples of erupted material from selected volcanoes under in-situ eruption conditions in order to mineralogically and petrophysically model the current processes. The reliability of the models will be examined by comparison of different volcano types. Visual video monitoring will provide direct correlation to the data sets observed at the Krakatau MP station. Sampling will start at the end of 2004. The integrated results shall contribute to a better risk assessment in the study area and to establishing improved early warning systems by definition of critical parameters. The project is funded by the German Federal Ministry of Education and Research (03G0578A).
V11B-1432 0800h
SPATIAL AND TEMPORAL CHANGES OF CODA Q IN THE GALERAS VOLCANO, COLOMBIA
Spatial and temporal changes of coda Q was analyzed for the Galeras volcano, located in the Andes of southwestern Colombia. The $Q^{-1}$ value was calculated using the Single Backscattering (SBS) working with seismograms filtered in frequency bands centered in 1.5, 3.0, 6.0 and 12.0 Hz. The temporal variation analysis was done following the frequency dependence $Q=Qof^{-n}$ before and after some eruptions. The temporal change of attenuation has two stages, the first is the 1989-1992 period, when the $Q^{-1}$ value shows an increasing tendency. The other is the 1992-2002 period that is characterized for a decreasing in the $Q^{-1}$ value. The rise of attenuation before July 92 eruption is associated with an andesitic dome extrusion, after this eruption the dome is destroyed and the $Q^{-1}$ value falls. Since 1994 low attenuation values are associated with the low activity stage of Galeras until recently activity at 2004. The frecuencial dependence shows that the $Q^{-1}$ value rises before an eruption and falls after this. The results show a clear change of attenuation before and after the eruptions so that coda $Q^{-1}$ is a premonitory tool for this volcano and we suggest it to be used in the Galeras activity monitoring. On the other hand, we have extended the first scattered concept for the inversion of spatial values of coda-Q. Two important anomalies was found, the first one at 9 km depth which is related to the main magmatic camera where are occurring process that feed the second one at 4km depth. Other minor anomalies has been correlated with faults that crossing the volcano.
V11B-1433 0800h
A Study of Cyclic Variation in the Properties of the Shallow Conduit Melt at Stromboli Volcano in
Stromboli volcano (Aeolian Islands, Italy) is generally considered a steady-state volcano among scientists and the public, with the current pattern of activity dating to at least the 7th century A.D. This reputation conceals the fact that typical activity at Stromboli fluctuates among ashing/puffing, weak explosions, and moderate/high intensity Strombolian explosions on time scales spanning minutes to months. Although Stromboli volcano has been intensely studied in recent years, the cause of such fluctuation remains largely unexplained. Our study compliments current knowledge by considering that evolving properties of the shallow melt phase are likely to influence eruptive activity. We present data on the microtextural characteristics of fresh ejecta from Stromboli volcano during 3 phases of variable eruption intensity in 2002. Activity shifted from mild, dominantly ash-charged explosions (January-May) into a phase of intense explosions (June-December), followed by an abrupt onset of the 2002/2003 effusive eruption (28 December). Our sets of lapilli from May, September/October, and December track changes in the shallow melt over this period. The observed textures indicate that this melt was a dynamic mixture of varying proportions of actively vesiculating magma, and stagnant, partly outgassed magma. Explosions in Sept/Oct and Dec ejected clasts composed of either type of melt independently, in addition to clasts formed from fine-scale mingling of both the actively vesiculating and stagnant magmas. In May only stagnant melt was ejected. We present observational and quantitative (vesicle size, chemistry, proportions) data for the ejecta, which highlight changes in properties of the shallow melt with time. We link this information with the variable explosive patterns witnessed at Stromboli in 2002.
V11B-1434 0800h
Duration-amplitude distribution of volcanic tremor at Soufri\`{e}re Hills Volcano, Montserrat
The duration amplitude distribution is examined for volcanic tremor occurring at Soufriere Hills Volcano, Montserrat. A period of tremor data from 2004 is considered. On the 3rd March 2004 there was a small collapse and explosion event lasting about 10$\,\rm$min. Low level continuous tremor started on the 10th March, increasing on the 15th March when it was accompanied with vigorous ash venting which continued for three days. The tremor continued at moderate to high levels with some banding until the 3rd April when the first spasmodic tremor was seen. The spasmodic tremor changes amplitude very rapidly and occurred for periods of about 24 hrs interspersed with periods of more continuous episodes, until 7th May when the tremor died away. The RSAM amplitudes derived from the broadband seismic records are used to analyse the distributions of the amplitude. During the two month period the duration-amplitude distribution is observed to change from an exponential distribution to a normal distribution. The exponential distribution results from a scale bound source for which either the driving force or source geometry is constant. For a normal distribution both the source geometry and the driving force must be constant. These results provide a constraint on models of source mechanisms of tremor.
V11B-1435 0800h
The effects of heat flow on low-frequency events
At many volcanoes the frequency content and waveforms of low-frequency seismicity is observed to vary over different timescales. At Soufriere Hills Volcano, Montserrat, single hybrid events merge into tremor and occasionally gliding lines are observed in the spectra indicating changes in the seismic parameters with time or varying triggering rates of single events. At Redoubt Volcano, Alaska more gradual variations are seen over longer timescales when the waveforms of a swarm of long-period events evolved gradually during a 23$\,\rm$hr time period. We consider the influence of heatflow on the seismic signal. Simple analytical solutions for heat flow are used to estimate the temperature gradient between the magma and country rock at increasing time intervals after the initial emplacement of magma. The seismic parameters and density are modelled by applying a tapering function of a similar shape to the numerical solutions between the seismic properties in the rock and those in the magma. The seismic wavefield for the conduit is modelled using a finite difference code. The effects of the taper on the seismic signal are considered for a range of acoustic parameters and timescales. The results show the long term variations that could be caused by heat flow and therefore facilitate the interpretation of variations in of the waveforms of low-frequency events.