Complex behavior and source model of the tremor at Arenal volcano, Costa Rica (original) (raw)
Related papers
Journal of Volcanology and Geothermal Research, 2006
Typical records of volcanic tremor and explosion quakes at Arenal volcano are analyzed with a high-resolution time-frequency method. The main characteristics of these seismic signals are: (1) numerous regularly spaced spectral peaks including both odd and even overtones; (2) frequency gliding in the range [0.9-2] Hz of the fundamental peak; (3) frequency jumps with either positive or negative increments; (4) tremor episodes with two simultaneous systems of spectral peaks affected by independent frequency gliding; (5) progressive transitions between spasmodic tremor and harmonic tremor; (6) lack of clear and systematic relationship between the occurrence of explosions and tremor. Some examples of alternation between two states of oscillation characterized by different fundamental frequencies are also observed. Some tremor and explosion codas are characterized by acoustic and seismic waves with identical spectral content and frequency gliding, which suggests a common excitation process. We propose a source model for the tremor at Arenal in which intermittent gas flow through fractures produces repetitive pressure pulses. The repeating period of the pulses is stabilized by a feedback mechanism associated with standing or traveling waves in the magmatic conduit. The pressure pulses generate acoustic waves in the atmosphere and act as excitation of the interface waves in the conduit. When the repeating period of the pulses is stable enough, they produce regularly spaced spectral peaks by the Dirac comb effect and hence harmonic tremor. When the period stability is lost, because of failures in the feedback mechanism, the tremor becomes spasmodic. The proposed source model of tremor is similar to the sound emission process of a clarinet. Fractures in the solid or viscous layer capping the lava pool in the crater act as the clarinet reed, and the conduit filled with low velocity bubbly magma is equivalent to the pipe of the musical instrument. The frequency gliding is related to variations of the pressure in the conduit, which modify the gas fraction, the wave velocity and, possibly, the length of the resonator. Moreover, several observations suggest that two seismic sources, associated with two magmatic conduits, are active in Arenal volcano. They could explain in particular the apparent independence of tremor and explosions and the episodes of tremor displaying two simultaneous systems of spectral peaks.
Geophysical Research Letters, 1997
Broadband seismic data recorded 2.3 km from the active vent of Arenal Volcano, Costa Rica, provide new constraints on tremor source processes. Arenal's tremor contains as many as seven harmonics, whose frequencies vary temporally by up to 75 percent, from initial values of 1.9 Hz for the first peak immediately following explosive eruptions to 3.2-3.5 Hz several minutes later. We infer that gas bubble concentration is variable within the conduit and also changes as a function of time, thereby changing the acoustic velocity. We infer that the source is a shallow, 200-660 m-long, vertically oriented 1-D resonator with matched boundary conditions, radiating seismic energy from a displacement antinode. Polarization analyses show that particle motion azimuths abruptly rotate, which may be explained by a decrease of the incidence angle. We suggest that energy is radiated predominantly from a displacement antinode that is changing position with time. Tremor consists mainly of transverse waves that travel at speeds of about 800 m/s. P waves in the magma conduit will couple very efficiently into S waves in the surrounding medium when there is virtually no impedance contrast between the two media for these two types of waves. The tremor at Arenal is similar to tremor at nine other volcanoes.
Non-linear explosion tremor at Sangay, Volcano, Ecuador
Journal of Volcanology and Geothermal Research, 2008
A detailed analysis of discrete degassing pulses, chugs, at Sangay volcano, was performed on seismic and infrasonic records to 8 determine the physics of the conduit. Infrasonic chugging signals appear as repetitive pulses with small variations in amplitude and 9 time lag. An automated time-domain analysis was developed to measure with high precision time intervals and amplitudes at 10 different wave arrivals, reducing the possibility error associated with hand picking. Using this automated method, a strong positive 11 correlation of acoustic amplitude with repose time between individual pulses on chugging signals of Sangay was found on 12 numerous oscillating sequences. Frequency gliding of apparent harmonic frequencies generally trends from high to low frequency 13 at Sangay, in contrast to trends at Karymsky Volcano, Russia. A new description of chugging events using wavelet transform 14 methods, appropriate for non-stationary signals, shows subtle changes in the waveforms relate to physical processes in the volcano. 15 A system of non-linear feedback, based on choked flow at the vent, is postulated as the most likely source of this volcanic tremor. 16 19 20 39 This maybe due to the fact that chugging usually follows 40 a larger Strombolian style explosion and pulsations 41 that occur in the aftermath are obstructed by the larger 42 amounts of gas and ash remaining from the initial blast. 43 The individual infrasonic chugging signals appear to be 44 discrete and time limited, often evolving over the length 45 of the chugging sequence. Corresponding seismic sig-46 nals, distorted because they are convolved with the in-47 tervening earth structure, display more complex signals
Journal of Geophysical Research-Solid Earth, 2014
We use wavefield decomposition methods (time domain cross correlation and frequency domain multiple-signal classification) to analyze seismic data recorded by a dense, small-aperture array located 2 km West of Arenal volcano, Costa Rica, and operated during 2.5 days. The recorded wavefield is dominated by harmonic tremor and includes also spasmodic tremor and long-period (LP) events. We find that the initial stages of LP events are characterized by three different wave arrivals. These arrivals propagate with similar back azimuths pointing to the volcano summit (∼80 • N) and increasing apparent slowness of 0.4, 1.1, and 1.7 s/km. Spasmodic tremors cannot be regarded as coherent signals. On the contrary, harmonic tremors are highly coherent, characterized by the stability of the apparent slowness vector estimates. Apparent slowness lays in the range 1-2 s/km. Back azimuths point in the general direction of the volcano but with a large variability (40-120 • N). Nevertheless, there are long-term variations and evidences of multiple simultaneous components in the harmonic tremor wavefield. These observations suggest that LP events and tremor are generated in a shallow source area near the volcano summit, although they do not share exactly the same source region or source processes. The tremor source is located in the shallowest part of the plumbing system, beneath the lava crust. This dynamic region is subject to complex fluctuations of the physical conditions. Degassing events at different locations of this region might generate variable seismic radiation patterns. The effects of topography and heterogeneous shallow structure of the volcano may amplify these variations and produce the wide directional span observed for volcanic tremor. On the other hand, the LP source seems to be more repeatable. LP events are likely triggered by fragmentation of the fluid flow in a slightly deeper portion of the volcanic conduits.
Bulletin of Volcanology, 2003
Following an initial phreatic eruption on 21 December 1994, activity at Popocatepetl has been dominated by fumarolic emissions interspersed with more energetic emissions of ashes and gases. A phase of repetitive dome-building and dome-destroying episodes began in March 1996 and is still ongoing at present. We describe the long-period (LP) seismicity accompanying eruptive activity at Popocatepetl from December 1994 through May 2000, using data from a three-component broadband seismometer located 5 km from the summit crater. The broadband records display a variety of signals, with periods ranging in the band 0.04-90 s. Long-period events and tremor with typical dominant periods in the range 0.3-2.0 s are the most characteristic signals observed at Popocatepetl. These signals appear to reflect volumetric sources driven by pressure fluctuations associated with the unsteady transport of gases beneath the crater. Very-long-period (VLP) signals are also observed in association with LP events and tremor. The VLP signals which accompany LP events display Ricker-like wavelets with periods near 36 s, whereas VLP signals associated with tremor waveforms typically show sustained oscillations at periods ranging up to 90 s. The spectra and particle motion patterns remain similar from event to event for the majority of LP and tremor signals analyzed during the time span of this study, suggesting a repeated, non-destructive activation of a common source. Hypocenters determined by phase pick analyses of selected LP events recorded by the seven-station, permanent Popocatepetl short-period network suggest that the majority of these events are confined to a source region in the top 1.5 km below the crater floor. The repetitive occurrences of VLP signals with closely matched waveform characteristics are consistent with a non-destructive reactivation of at least two sources. One source appears to coincide with the main source region of LP seismicity, whereas the second is a deeper source whose activity appears to be intimately linked with episodes of monochromatic tremor.
SLOW WAVES TRAPPED IN A FLUID-FILLED INFINITE CRACK: IMPLICATION FOR VOLCANIC TREMOR
The dynamics and seismic radiation of fluid-filled cracks have been studied by numerous authors, as models for tremor and for long-period events observed at volcanoes. One of the most intriguing results of the recent models is the existence of a very slow wave propagating along the crack boundary. In order to better understand this slow wave, which has so far only been studied numerically, we studied analytically normal modes trapped in a liquid layer sandwiched between two solid half--space,;. A slow wave, similar to the tube wave found by Blot, exists for all wavelengths. In the short wavelength limit, this wave approaches the Stoneley wave for the liquid-solid interface. Unlike the tube w; re, however, as the wavelength increases to infinity, both the phase and group velocities approach zero, in inverse proportion to the square root of wavelength. The phase velocity and amplitude of this slow wave are in good agreement with those obtained by the numerical studies on the dynamics of fluid-filled cracks by two-dimensional and three-dimensional finite difference methods. In the past the size of a magma body has been estimated from volcanic tremor periods and the acoustic velocity in the fluid. These estimates should be drastically reduced if the slow wave donfinates (he tremor. For exan?ple, the extremely long-period volcanic tremor, with periods up to "s, observed at Mount Aso may be generated by a fluid-filled crack of modest size, a magma body 0.5 m thick end 0.5 km long.
Source and path effects in the wave fields of tremor and explosions at Stromboli Volcano, Italy
Journal of Geophysical Research, 1997
The wave fields generated by Strombolian activity are investigated using data from small-aperture seismic arrays deployed on the north flank of Stromboli and data from seismic and pressure transducers set up near the summit crater. Measurements of slowness and azimuth as a function of time clearly indicate that the sources of tremor and explosions are located beneath the summit crater at depths shallower than 200 rn with occasional bursts of energy originating from sources extending to a depth of 3 km. Slowness, azimuth, and particle motion measurements reveal a complex composition of body and surface waves associated with topography, structure, and source properties. Body waves originating at depths shallower than 200 rn dominate the wave field at frequencies of 0.5-2.5 Hz, and surface waves generated by the surficial part of the source and by scattering sources distributed around the island dominate at frequencies above 2.5 Hz. The records of tremor and explosions are both dominated by SH motion. Far-field records from explosions start with radial motion, and near-field records from those events show dominantly horizontal motion and often start with a low-frequency (1-2 Hz) precursor characterized by elliptical particle motion, followed within a few seconds by a highfrequency radial phase (1-10 Hz) accompanying the eruption of pyroclastics. The dominant component of the near-and far-field particle motions from explosions, and the timing of air and body wave phases observed in the near field, are consistent with a gaspiston mechanism operating on a shallow (<200 rn deep), vertical crack-like conduit. Models of a degassing fluid column suggest that noise emissions originating in the collective oscillations of bubbles ascending in the magma conduit may provide an adequate self-excitation mechanism for sustained tremor generation at Stromboli. 15,129 15,130 CHOUET ET AL.: SOURCE AND PATH EFFECTS AT STROMBOLI VOLCANO 15 ø 10.28' 38 ø 49.49' 38 ø 45.98' 15 ø 14.86' GINO I o e o ß ß ß o ß o ß ß ß ß o ß o lOO 2oo M
Permanent tremor of Masaya Volcano, Nicaragua: Wave field analysis and source location
Journal of Geophysical Research, 1997
The Masaya Volcano, Nicaragua, is a basaltic caldera in a subduction zone. The permanent source of the volcanic tremor was located inside Santiago crater, at the lava lake's position and 400 m below the NE rim, and therefore corresponds to superficial magma activity. We used two tripartite arrays (90 m side), one semicircular array (r=120 m) in 1992, and two semicircular arrays (r=60 m) and a 2500 m long linear array radiating out from the source and on the flank of the crater in 1993. We used both a cross-spectrum method and a correlation method to determine the wave delay time between the reference station and the other stations of an array and to quantify the wave field. Using the delays therefore by intersecting the back azimuth wave directions from the arrays, we could pinpoint the source. Additionally, the correlation coefficients obtained as functions of frequency for the three components of motion confirm the inferred position of the source of tremor. The tremor's wave field is composed of comparable quantities of dispersed Rayleigh and Love surface waves, whose phase velocities lie in the ranges 730-1240 m/s at 2 Hz and 330-550 m/s at 6 Hz. The dispersive phase velocities were inverted to obtain crustal structures with a minimal number of layers. The resulting velocity models are similar for the northern and southern parts of the volcano. After geometrical spreading corrections, Q2Hz = 14 and Q3Hz =31 were determined along the northern linear array. The typical low velocities and low Q corresponding to the cone structure and are similar to those of other basaltic volcanoes like Puu Oo, Hawaii, and Klyuchevskoy, Kamchatka.
Journal of Volcanology and Geothermal Research, 2003
Villarrica volcano (Chile, 39.42 ‡S, 71.93 ‡W) is one of the most active volcanoes in the Andes. It shows persistent strombolian activity in a small lava lake situated at the bottom of the summit crater. From September to December 2000, the volcano exhibited anomalous behaviour: after a regional tectonic seismic event on 22 September, the eruption of a small volume of lava apparently plugged the conduit. This situation then lasted for ten days. On 5 and 8 October, a series of explosions reopened the system. This eruption behaviour has allowed us to study the evolution of the characteristic tremor in the open and closed states of the conduit: a shift in the peak frequency from 1 Hz (open) to 2 Hz (closed). The explosions related to the opening of the conduit were forecasted by means of a material failure method.