observations of volcanic tremor during the january ... · recorded several episodes of low...
TRANSCRIPT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Observations of volcanic tremor during the January-February 2005 eruption of Mt.
Veniaminof, Alaska.
Silvio De Angelis and Stephen R. McNutt
Alaska Volcano Observatory – Geophysical Institute, University of Alaska Fairbanks, 903 Koyukuk Drive PO
BOX 757320, Fairbanks, Alaska, 99775-7320, USA.
Prepared for submittal to Bullettin of Volcanology
Contact author: Silvio De Angelis
e-mail:[email protected];
phone +1-907-474-7234
1
Abstract 25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Mt. Veniaminof, Alaska Peninsula, is a strato-volcano with a summit ice-filled caldera
containing a small intracaldera cone and active vent. From January 2 to February 21, 2005,
Mt. Veniaminof erupted. The eruption was characterized by numerous small ash emissions
(VEI 0 to 1) and accompanied by low-frequency earthquake activity and volcanic tremor. We
have performed spectral analyses of the seismic signals in order to characterize them and to
constrain their source. Continuous tremor has durations of minutes to hours with dominant
energy in the band 0.5-4.0 Hz, and spectra characterized by narrow peaks either irregularly
(non-harmonic tremor) or regularly spaced (harmonic tremor). The spectra of non-harmonic
tremor resemble those of low-frequency events recorded simoultaneously to surface ash
explosions, suggesting that the source mechanisms might be similar or related. We propose
that non-harmonic tremor at Mt. Veniaminof results from the coalescence of gas bubbles and
low-frequency events are related to the disruption of large gas pockets within the conduit.
Harmonic tremor, that is characterized by regular and quasi-sinusoidal waveforms, has
duration of hours. Spectra, containing up to five harmonics, suggest the presence of a
resonating source volume that vibrates in a longitudinal acoustic mode. An interesting feature
of harmonic tremor is that frequency is observed to change over time: spectral lines move
towards higher or lower values while the harmonic nature of the spectra is maintained. Factors
controlling the variable characteristics of harmonic tremor include changes in acoustic
velocity at the source and variations of the effective size of the resonator.
Keywords: Volcanic tremor, harmonic tremor, low-frequency events, conduit resonance,
ash eruptions, Mt. Veniaminof, volcanic seismology
Introduction
2
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
Over the past decade the observation of low-frequency (LF) earthquake activity at volcanoes
has become increasingly important in monitoring and forecasting eruptions. LF signals with
durations of minutes to days or longer, frequently observed near active volcanoes, are usually
referred to as volcanic tremor (Aki and Koyanagi, 1981; Fujita et al., 1995; Hellweg, 2000;
McNutt, 2002); tremor has been documented at 160 world volcanic centers (McNutt, 1994)
and its detection is an important part of most volcano monitoring programs.
The most general appearance of a tremor waveform is that of a continuous signal with
emergent onset, smoothly varying amplitudes and energy confined in the band 1.0-5.0 Hz
(Julian, 1994; Fujita et al., 1995; Neuberg et al., 2000; McNutt, 2002). Tremor spectra can be
either broadband without pronounced peaks, or characterized by a variable number of
regularly spaced peaks. If the spectrum contains either a single peak, or a variable number of
regularly spaced peaks, the signal is called harmonic tremor.
Unlike tectonic earthquakes, that involve mechanisms of shear failure of rock at the source,
tremor originates from complex fluid-rock interactions within volcanoes. While volcanic
tremor is a common precursor and accompanies most volcanic eruptions, its characteristics
including depth, duration and amplitude, can vary considerably. The broad range of tremor
properties suggests that multiple mechanisms may be responsible for its generation, even at
the same volcano; several models have been proposed in order to account for tremor
generation including free oscillations of fluid filled cavities (Sassa, 1936; Crosson and Bame,
1985; Fujita et al., 1995), jerky crack propagation (Aki et al., 1977), flow-induced oscillations
of volcanic conduits (Julian, 1994; 2000; Hellweg, 2000), and the resonance of fluid filled
cracks and conduits (Chouet, 1987; Benoit and McNutt, 1997; Garces, 1997; Garces and
McNutt, 1997). Earlier models of tremor, based on the free oscillations of magma chambers,
were able to reproduce peaked harmonic spectra but often relied on unrealistic dimensions of
3
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
the resonating volumes. Most recently, the study of tremor has received increasing attention
by volcano-seismologists because of its potential as a monitoring and forecasting tool for
unrest at volcanoes, and more refined models have been proposed. Hellweg (2000) suggested
that the presence of numerous overtones in harmonic spectra, and their exact relationship to a
fundamental frequency, is the result of non-linear flow conditions in pipe-like conduits;
turbulence in conduit flows with high Reynolds numbers, may generate periodic pressure
disturbances and produce regularly peaked spectra characteristic of harmonic tremor. Julian
(1994; 2000) proposed that tremor results from the oscillations of slot-like channels with
movable elastic (damped) walls, induced by the flow of a viscous incompressible fluid. This
model is described by a 3rd order system of non-linear differential equations whose solutions
are controlled by the fluid flow pressure; increasing values of this parameter will account for
steady flow without oscillations, short-lasting oscillations, sustained oscillations, period-
doubling cascades, and chaotic oscillations controlled by non-linear attractors. Gordeev (1993)
and Schlindwein et al. (1995), showed that peaked harmonic spectra can be reproduced by the
convolution of a series of equi-spaced spikes, i.e., a Dirac comb funcion with a source
wavelet. The convolution of an arbitrary function with a Dirac comb yields a series of replicas
of the original function with period ΔT, equal to the spacing of the teeth of the comb. The
theoretical spectrum of the signal consists of a fundamental frequency (defined by 1/ΔT) along
with a number of integer overtones, and is modulated by the spectrum of the source function.
The generation of volcanic tremor by resonating fluid filled fractures has been extensively
treated in the literature, as well. Aki et al. (1977) proposed that tremor is generated by the
pressure driven motion of fluids through a chain of cracks connected by narrow channels; the
characteristics of tremor are controlled by parameters such as the length of the cracks and the
4
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
fluid pressure. Chouet (1987) suggested that the resonance modes of a fluid-filled rectangular
crack, triggered by a localized pressure disturbance (acting on the crack walls), correspond to
the peaks observed in the spectra of long period earthquakes and volcanic tremor. The
predicted wavefield depends on parameters that include the crack dimensions, the position and
intensity of the pressure disturbance and the impedance contrast between the fluid and the
surrounding rocks. McNutt (1986) and Benoit and McNutt (1997) modeled the source of
harmonic tremor as a 1D vertical resonating conduit filled with gas-charged magma, each of
the observed spectral peaks representing an eigen-mode of vibration of the oscillator. Mori et
al. (1989) explained the observations of harmonic tremor at Langila volcano, Papua New
Guinea in terms of a gas filled resonating volume. The resonance modes of 1D oscillators are
controlled by the length of the resonator, the acoustic properties of the fluid, and a set of
specified boundary conditions.
In this paper we will present observations and spectral analyses of volcanic tremor recorded
during the January-February 2005 eruption of Mt. Veniaminof, Alaska. This eruption, has
been well documented with seismic, satellite and web camera observations, and provides the
best characterization to date of the eruption style of Mt. Veniaminof volcano.
Background
Mt. Veniaminof is a large stratovolcano on the Alaska Peninsula (56.2° N, 159.4° W,
elevation: 2507 m), 35 km wide at the base, truncated by a steep-walled caldera 8x11 km in
diameter that formed about 3700 years B.P. The caldera is filled by an ice field that ranges in
elevation from 1750 to 2000 m; an intra-caldera cone is located in the western part of the
caldera with a small summit crater. The cone has an elevation of 2156 m, about 330 m above
the surrounding ice field (Miller et al., 1998) and is the site of all historical eruptions (Simkin
5
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
and Siebert, 1994). A belt of Quaternary cinder cones (Detterman et al., 1981a,b) extends in
the SE-NW direction from the main volcanic edifice to the Bering Sea coast.
The recent activity includes moderate Strombolian eruptions in 1983, 1993 and 1994 from the
intra-caldera cone accompanied by lava flows and, mild explosive activity in 2002 and 2004.
Ash and steam explosions are a characteristic feature of the eruptive activity at Mt.
Veniaminof.
In the summer of 2001, the Alaska Volcano Obsevatory (AVO) installed a network of 8
vertical component, short-period (1s) seismometers (Mark Products L4-C) at Mt. Veniaminof
(Figure 1); continuous analog data are transmitted via radio telemetry and phone lines to AVO
offices where they are digitally recorded at a sample rate of 100 Hz (Thompson et al., 2002).
Archiving of data began in February 2002 and, since then, different types of signals have been
recorded including LF earthquakes, volcano tectonic (VT) earthquakes, and volcanic tremor.
While low rates of VT activity (< 1 event/day) constitutes the seismic background at Mt
Veniaminof during quiescent periods, the occurrence of a large number of LF events and
volcanic tremor characterized the eruptions in 2002 and 2004 (Sanchez, 2005; Sanchez et al.
2005). During the eruptions in January-February 2005, up to 20-30 LF events/hour were
detected. In addition to the seismic network, a web camera located in the town of Perryville,
about 35 km from the volcano, provides a photographic record of the surface activity
(photographs taken every 5 minutes). Images from the Advanced Very High Resolution
Radiometer (AVHRR) sensor on the NOAA-12 and NOAA-14 satellites are also used to
detect volcanic eruption clouds and thermal anomalies.
6
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
Eruption chronology
After a relatively long period of quiescence of about 6 years, Mt. Veniaminof first showed
signs of unrest in 2002-2004. During September 2002, pulses of volcanic tremor were detected
and bursts of steam (possibly containing small amounts of ash) were observed rising above the
intra-caldera cone; intermittent steaming activity accompanied by variable seismicity
continued until April 2003 (Sanchez, 2005). In the period February-October 2004, AVO
recorded several episodes of low amplitude volcanic tremor and small low-frequency volcanic
earthquakes accompanying ash and steam emissions from the intra-caldera cone. Short
duration episodes of harmonic tremor were recorded on October 12 and November 2, 2004.
During November and December 2004 no significant surface activity was observed and the
seismicity remained at background levels. On January 1, 2005, AVO started to record weak
tremor that lasted for about two days. On January 4, the signal character changed to that of
numerous low-frequency earthquakes (about 1-2 per minute), and the level-of-concern color
code (Table 1) was upgraded from green to yellow. Ash outbursts were observed rising to
heights of few hundreds meters above the intra-caldera cone starting January 3. The following
week was characterized by elevated levels of seismicity, and the episodic surface activity
evolved into more continuous emissions forming ash plumes rising up to about 1300 m above
the vent (Figure 2). Starting January 7, satellite images showed a persistent thermal anomaly
in the vicinity of the active cone; both seismic and surface activity exceeded the levels
observed during 2002-2004. On January 10, the level-of-concern color code was upgraded to
orange. Amplitude and occurrence of volcanic tremor and LF events increased over the month
of January, and the activity at the surface consisted of emissions forming ash clouds and ash
fall reaching outside the caldera boundary. On February 4, incandescence was clearly visible
7
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
on the night-time web camera images, indicating strombolian activity with ejection of hot
blocks from the intra-caldera cone. This activity was accompanied by an increase in the
amplitude and occurrence rate of LF earthquakes and continuous low amplitude volcanic
tremor. From February 5 onward, the seismicity was dominated by volcanic tremor; the
activity ranged from tremor bursts with durations of few tens of seconds to continuous tremor
lasting hours. Harmonic tremor was mostly observed between February 17 and February 21.
The pattern of seismicity was consistent with mild explosive activity from the intra-caldera
cone, although Mt. Veniaminof was not visible in the web camera images for almost the entire
month of February due to cloudy weather. On February 21, harmonic tremor was sporadically
observed along with non-harmonic tremor and LF events; earthquake activity abruptly ended
late in the evening the same day, marking the end of the eruption. The level-of-concern color
code was downgraded to yellow on February 25 and to green on March 4, 2005. Table 1
summarizes the observations during the course of the eruption.
Data analyses and results
The seismic activity recorded at Mt. Veniaminof during January-February 2005 included LF
earthquakes, non-harmonic and harmonic tremor. The occurrence of many LF earthquakes
overlying a low level non-harmonic tremor signal, characterized the eruption during January
and at the beginning of February 2005. Moderate to strong pulses of tremor (durations of few
tens of seconds to minutes) superimposed on a low-amplitude continuous signal, as well as
strong harmonic tremor lasting up to hours, were mostly observed during the month of
February. We performed spectral analyses of about two months of tremor data in order to
characterize the properties of the signal over the duration of the eruption and to constrain its
8
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
source. All data were band-pass filtered between 0.4 and 10.0 Hz in order to reduce the effects
of ocean microseisms and incoherent higher frequency noise.
Seismograms and amplitude spectra of non-harmonic tremor recorded at station VNSS (5.4
km from the vent) are shown in Figure 3; energy is confined to the band 0.5-5.0 Hz and
spectra are characterized by narrow and irregularly distributed peaks. It is worth to note that
the seismic sensors installed at Mt. Veniaminof have a natural period of 1 second, and they
may not allow to recognize spectral peaks at very low frequencies. For each trace, the
amplitude spectrum was averaged by use of the direct segment method (Bath, 1974), stacking
the amplitude spectra of six adjacent windows of data (10 seconds duration each). The use of
this technique enhances the part of the spectra common to the entire waveform while reducing
the contributions of incoherent noise. The stacked spectra (fig. 3a-c), exhibit an overall broad
triangular shape peaked between 1.0 and 2.0 Hz. We compared the frequency content of non-
harmonic tremor with that of low-frequency earthquakes recorded during periods of ash
explosions activity at the volcano. Figure 4 is an example of velocity seismogram and
amplitude spectrum of a low frequency event recorded at station VNNS; energy appears
concentrated in the 1.0-4.0 Hz band with dominant peaks between 1.0 and 2.0 Hz.
Figure 5 shows six consecutive hours of harmonic tremor recorded at station VNSS on
February 18, 2005. Beneath each waveform the corresponding spectrogram is shown. The
spectrograms were generated moving a 1024 sample (10.24 s at 100 Hz sampling rate) sliding
window along the waveform, and calculating its periodogram for 512-sample overlapping
positions of the window. Energy is spread over the interval 0.5-5.0 Hz; the fundamental
frequency, in the band 0.5-2.0 Hz, is accompanied by one to three integer harmonics. Figure 6
shows velocity seismograms and the amplitude spectra of one-minute samples of harmonic
9
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
tremor recorded at station VNSS; the spectra are characterized by sets of integer harmonics
which are multiples of the 1.1 Hz fundamental frequency.
More complicated spectra, containing up to five harmonics, were occasionally observed. For
instance, figure 7 shows 1 minute of harmonic tremor recorded at station VNSS and its
amplitude spectrum; a set of 5 regularly spaced peaks that are integer multiples of 0.6 Hz, is
clearly visible.
The frequency content of harmonic tremor was checked for consistency across the seismic
network. In Figure 8 we show the velocity seismograms and power spectral density of 1
minute data samples recorded at four different sites; the fundamental mode and the first
harmonic appear clearly in the spectra at all stations suggesting that it may reflect some source
characteristic. On the other hand, as frequency increases, spectral peaks become less visible at
the stations located at greater distances from the active vent; this is, probably, the result of
increased attenuation that affects higher frequency waves as they travel away from the source.
Indeed, seismic data at Mt. Veniaminof are recorded by instruments located at distances
between 10.8 and 20.4 km from the active cone, the only exception being station VNSS that is
located 5.4 km from the vent.
An interesting feature of harmonic tremor at Mt. Veniaminof is that frequencies are observed
to change over time; spectral lines systematically move towards lower (converging lines) or
higher (diverging lines) values while the harmonic structure of the signal is maintained. This
time dependent feature of tremor, known as spectral gliding, has been observed at a number of
other world volcanoes including Mt. Semeru, Indonesia (Schlindwein et al, 1995), Arenal,
Costa Rica (Benoit and McNutt, 1997; Hagerty et al,, 2000; Julian, 2000), Lascar, Chile
(Hellweg, 2000), and Montserrat, West Indies (Neuberg et al., 2000). Gliding episodes at Mt.
Veniaminof include frequency changes between 30% and 75 % with respect to the initial value
10
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
over periods of 2-3 minutes to 10-15 minutes. Figure 9 shows about 50 minutes of harmonic
tremor with variable harmonic frequencies; within a time interval of about 25 minutes, three
gliding episodes (t1, t2, t3) can be distinguished. During period t1, the fundamental frequency
decreases from 1.8 Hz to 0.9 Hz in about 300 s; period t2 shows an increase from 1.0 Hz to 1.6
Hz in 800 s; during t3 there is a decline from 1.7 Hz to 1.1 Hz in 340 s.
Discussion
We have reviewed data for about two months of seismic activity at Mt. Veniaminof and
performed spectral analyses of volcanic tremor and LF earthquakes occurring during the
January-February 2005 eruption. Two types of volcanic tremor were identified: broadband
non-harmonic and harmonic. The spectral analyses of non-harmonic tremor and low-
frequency earthquakes recorded during periods of mild Strombolian activity demonstrate a
notable similarity between the two types of seismicity, suggesting that they may share similar
sources. In analogy to the mechanism proposed by Ripepe and Gordeev, (1999), we suggest
that non-harmonic tremor is produced by coalescence (free or forced) of gas bubbles from a
layer of smaller bubbles, through the surface of a magma column. LF events, accompanying
the explosive emission of ash at the surface, are likely related to the explosive disruption of
individual large gas pockets within the magma in the volcanic conduit.
Harmonic tremor mostly occurred during the last stages of the eruption; durations of minutes
to hours and spectra consisting of one up to five harmonics in the band 0.5-5.0 Hz were
typical. Harmonic frequencies were observed to change over time, by up to 75% with respect
to the initial value in 5 minutes. While the irregularly peaked spectra of non-harmonic tremor
and low-frequency events are attributed to mechanisms that involve gas buble coalescence and
gas explosions, the simple and regular spectral structure of harmonic tremor seem to require a
11
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
different interpretation in terms of its source mechanisms. In the introduction section of the
manuscript we discussed various theoretical models potentially able to account for the
generation and features of volcanic tremor. They included models based on the flow-induced
oscillation of elastic volcanic conduits and on the resonance of gas or fluid filled cavities.
Models based on flow-induced oscillations generally require long and narrow conduits and
high flow velocities. Hellweg (2000) showed that the generation of harmonic tremor by
turbulence in conduit flow, would require conduit length to diameter ratios greater than 50,
and flow velocities on the order of 100 m/s for andesitic magma; such magma flow velocities
over time periods as long as hours or days, are unlikely at a volcano that is not undergoing a
continuous and vigorous explosive eruption. Julian (1994, 2000) proposed a model of
incompressible Newtonian fluid flow through a slot-like fissure with elastic damped walls.
This mechanism involves flow pressures on the order of 10-15 Mpa and flow velocities of 50-
100 m/s, to produce frequencies on the higher end of tremor observed at Mt. Veniaminof
(about 5 Hz). The flow of an incompressible fluid with high flow velocities, seems not
appropriate to our case, although it may be suitable when more continuous and vigorous
eruptive activity is observed.
Models that involve the resonance of fluid filled cavities appear to be more suitable candidates
to explain the characteristics of the harmonic tremor observed at Mt. Veniaminof. A pressure
disturbance applied at the walls of a fluid filled fracture (Chouet, 1988) of length L and width
W, for example, can generate standing wave vibrations in the crack; two sets of resonance
modes, longitudinal and lateral (2L/n and 2W/n (n=1,2,3,…)) would be observed. We don’t
have evidence of 2D resonance modes, although these are likely to appear at frequencies
higher than those observed in the spectra at Mt. Veniaminof, that may have been attenuated
travelling from the source region to the distant receivers.
12
We infer that the features of harmonic tremor at Mt. Veniaminof are explained by a 1D
resonant source. In 1D resonating conduits, acoustic wavelengths excited by pressure
disturbances propagate along a gas or fluid column, and upon reflection in correspondence
with specific boundaries, interfere with the initial wave. For certain characteristic frequencies
of the traveling waves, resonance can be established and energy is radiated into the ground in
the form of seismic waves through coupling with the surrounding rocks. Specific boundary
conditions control the reflection in correspondence of the terminations of the resonant
structure, and its modes of vibration. A pipe-like conduit that is an open-open or closed-closed
system (matched boundary conditions) has
285
286
287
288
289
290
291
292
/ 2nλ (λ =wavelength, n=1,2,3,…) waves as
longitudinal resonance modes whereas a system with one open and one closed end (unmatched
boundary conditions) has (2
293
294
1) / 4n λ− (λ =wavelength, n=1,2,3,…) waves. In terms of
frequency, pipes with matched boundary conditions are characterized by spectra with evenly
spaced peaks consisting of the fundamental mode,
295
296
0f , and a set of integer harmonics which
are multiples of
297
0f ( 0 1 0 2 0 , ...f 0, 2 , 3 , ; 1,..,nf f f f nf n N= = = =298
299
f ); spectra that have equally
spaced peaks and contain only odd harmonics
( 0 1 0 2 0 0, 3 , 5 , ... , (2 1) ; 1,..,nf f f f f f n f n= = = − = N300
301
302
303
304
305
306
307
) characterize resonant systems with
unmatched boundary conditions. A practical realization of this model is a conduit filled with
gas or bubbly magma bounded at the bottom by more viscous and incompressible magma. The
interface between the two phases is likely to correspond to the nucleation level within the
conduit. Because of the strong impedance contrast between the source fluid and the underlying
non vesciculated magma, this termination acts like a closed termination. The upper end of the
conduit can be either open to the atmosphere and act as an open termination, or obstructed by
a relatively viscous plug at the vent acting as a closed boundary. However, as noted by Garces
13
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
and McNutt (1997), the observation that the conduit is plugged at the vent does not necessary
imply that it is an acoustically closed boundary and, similarly, an open vent does not necessary
behave as an open termination. In fact, a conduit that is plugged may enlarge at the vent,
causing this boundary to act as an open termination; on the other end an open conduit may act
as a closed boundary if it’s constricted at the vent.
As opposed to other volcanoes, we do not have evidence of significant emissions of fluid lava
at the surface neither continuous nor episodic, and the strength of tremor, measured through
reduced displacement, is small. Benoit and McNutt (1997), suggested that reduced
displacement of the order of 20 cm2 is typical for sources processes that involve gas charged
magma, while lower values, on the order of 5 cm2 or less, have been measured for sources
involving pure gas or hydrothermal fluids. Daily averages of reduced displacement at station
VNSS (the closest to the active vent) rarely exceeded 1.5 cm2 during the eruptive period.
On the basis of the available observation, we believe that the resonance of a volume filled with
gas bounded at its bottom by column of magma, may account for the features of harmonic
tremor at Mt. Veniaminof. Degassing from the underlying magma may set up standing waves
in the gas chamber, hence producing harmonic spectra. The spacing of spectral peaks observed
in Figure 6 suggests that the system has matched boundary conditions. Schlindwein et al.
(1995), also noted that the greater homogeneity of gases with respect to magma, allows the
development of long lasting oscillations without disturbing the harmonic nature of the signal
even when frequency changes may occur.
The frequency of the signal can be measured from its spectrum; by assuming specific acoustic
properties of the source mixture, the length of the resonator, L, can be estimated using the
following equation (Hagerty et al., 2000):
14
02vLf
= (1.1) 331
where v is the acoustic velocity of the gas/fluid at the source and 0f , is the fundamental
frequency measured in Hz. Mori et al. (1989) suggested that an upper limit for the acoustic
velocity of gas in volcanic conduits at shallow depths is about 300 m/s. Schlindwein et al.
(1995) have used a wave velocity in hot air of 500 m/s from Bergmann and Schaffer (1990).
The sound speed of pure gas mixtures can be calculated in the most simple way through the
formula:
332
333
334
335
336
337
( )gas gasc R Tγ= ⋅ ⋅ / M338 (1.2)
where γ is the is the adiabatic constant for the gas, T is the absolute temperature, Rgas is the
universal gas constant, and M is the molecular mass of the gas species. For gas compounds of
H2O, CO2, and SO2 , within a reasonable range of temperatures for volcanic systems (300-800
K), this formula produces estimates of sound velocities on the order of 102 meters per second.
The topic of the sound speed of fluids and gas mixtures, however, is complex; while single-
phase mixtures are relatively easy to treat, the sound velocity of multi-phase compounds, as
found in actual volcanic environments, is dramatically different from the that of either pure
component (De Angelis, 2006). In liquid-gas mixtures, that have the density of liquids but the
compressibility of gases, even very small variations of the volume fraction of gas can greatly
reduce the sound speed from a few hundred to a few meters per second (Kieffer, 1977).
Kumagai and Chouet (2000), have studied the acoustic properties of ash-gas mixtures; they
found that sound velocity for SO2–ash mixtures varied from 100 to 500 m/s depending on the
temperature, the size of solid particles, the pressure and the composition of the mixture. For
the purpose of a first order estimate of the linear dimension of the resonant conduit, we
339
340
341
342
343
344
345
346
347
348
349
350
351
352
15
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
assume a velocity of 300 m/s. Using this value, and considering the 1.1 Hz fundamental mode
in the spectra of Figure 6, a trivial mathematical exercise would give a conduit length of about
136 m according to equation (1.1). Although, we can’t constrain the sound velocity of the gas
mixture at the source with such a degree of accuracy, based on values widely used in literature
(300-500 m/s), the linear dimension of the resonant conduit can be constrained to be on the
order of 102 m. While, we favour a gas phase at the source, it still is not possible to discard
entirely the hypothesis that the source fluid is bubbly magma. If this was the case, the wave
velocity of volcanic fluids may vary from 2500 m/s (Murase and McBirney, 1973) to as low as
300 m/s (Aki et al., 1978) according to different flow conditions and magma properties (for
example, the more rich in gas the magma, the lower the wave velocity).
Spectra like those of figure 7 showing 5 well defined peaks, were occasionally observed. One
possible interpretation is a resonant conduit with matched boundary conditions; in this case, a
velocity of 300 m/s would require a 250 m-long resonating conduit to account for the observed
0.6 Hz standing wave vibration. Another option may be that the conduit at Mt. Veniaminof
generates resonant modes corresponding to both matched and unmatched conditions. Similar
observations have been already reported at Mt. Spurr by Garces and McNutt (1997). For short
periods of time, one of the conduit terminations may act as a partially open and partially
closed boundary (i.e. partial reflection and partial transmission); as an example, the vent may
be partly obstructed and partly open to the atmosphere. This model is attractive because 0.6
Hz is roughly half of the most commonly observed 1.1 Hz fundamental peak. Thus the main
structure may remain intact, and only the boundary conditions need to change to produce the
two sets of peaks.
The obvious gliding of spectral lines in figure 9 can be attributed to either variations of the
acoustic properties at the source or changes in the effective length of the resonator. Changes of
16
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
the acoustic velocity may results from a handful of factors such as mixing of different volatile
species (e.g. H2O, CO2, SO2) in variable proportions, or development of multi-phase mixtures
due to the presence of water and steam at shallow depths in the volcanic conduit. Variations in
the effective length of the resonator, instead, involve rising or falling of the bubbly magma
within the volcanic conduit. For instance, the gliding shown in figure 9 would require the
magma below the gas chamber to rise (upward gliding) or fall (downward gliding) at
velocities of about 0.2 m/s (t1), 0.08 m/s (t2) and 0.15 m/s (t3). These velocities are not
unreasonable for the flow of a bubbly magma within a volcano. Closer inspection of the
spectrogram in figure 9, reveals the presence of bumps and wiggles of the fundamental
spectral line, superimposed on the overall upward gliding. Frequency changes by about 0.1-0.2
Hz, roughly corresponding to 10-20 % of the most common fundamental frequency, over time
periods of few seconds. We think that this short term oscillatory behavior may result from
transient instabilities such as changes in density and composition of the volatile phases;
alternatively, it may be attributed to small and relatively rapid variations of the magma column
height around a position of unstable equilibrium caused, for instance, by quasi-periodic
collapse of gas bubbles (this would result in short term variations of the the effective length of
the resonator). Similar observations have been previously reported at Arenal volcano, Costa
Rica (Hagerty et al., 2000).
It is important to point out that resonance is not a “source mechanism” itself. Resonant
conduits define “source regions” that, through the propagation of standing and interface waves
excited by a “source trigger”, can generate harmonic spectra. The source trigger is a
mechanism that releases the elastic energy necessary to kickstart resonance. In the specific
case of Mt. Veniaminof we infer that the trigger mechanism is represented by degassing pulses
at the lower boundary of the resonator. The period of the signal is then stabilized by a
17
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
feedback mechanism related to the resonance itself. The long duration of the harmonic signals
suggests that the trigger should be continuous; this may be obtained through pressure
oscillations in the resonator that, once initiated by a degassing pulse from the underlying
magma, in turn control further degassing (Schlindwein et al., 1995).
We suggest that the seismic activity during November-December 2005 and early January 2006
at Mt. Veniaminof was related to the re-activation of the system and ascent of new magma
reaching up the surface. Non-harmonic volcanic tremor and low-frequency earthquake activity
in early January 2006, reflected degassing through the surface of a magma column that had
reached shallow depths. Larger ash explosions and Strombolian activity during the second half
of January and early February 2006 were accompanyied by swarms of LF events with larger
amplitudes than previously. We have suggested that these LF earthquake are related to the
explosive fragmentation of magma within the conduit. Between early and mid-February,2006
the column of magma may have slowly dropped at a lower level in the conduit, leaving a
“plug” of viscous material at the vent. Under these conditions harmonic tremor, may have
been generated around mid-February, 2006. On February 21, 2006 the earthquake activity at
Mt. Veniaminof abruptly ended. Since then only minor and episodic ash explosions have been
observed. It can be considered that the volcano has returned to its normal state of persistent
low-level seismic and surface activity. The cartoon in Figure 10, illustrates the evolution of
activity at Mt. Veniaminof volcano between late 2004 and March 2005.
Even though a large variety of models exist accounting for the occurrence and features of LF
seismic activity and volcanic tremor, their source mechanisms and propagation are still not
fully understood, and unanimous consensus has not been reached on these topics. Theoretical
models often rely on a number of assumptions. We are aware, for example, of the limitations
of the simple model of a fluid filled resonant cavity and that it can not ultimately explain the
18
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
manifold complexities of harmonic tremor. The intent of this manuscript, is to present
observations of seismic activity from an interesting case study and to point out the role that a
resonating source may have in the generation of harmonic tremor at Mt Veniaminof. Many
aspects need to be refined in the formulation of tremor models. Particularly relevant are the
determination of the stress conditions at fluid-solid boundaries within volcanic systems,
leading to a comprehesive understanding of the conditions under which acoustic energy and
pressure disturbances within conduits are transformed into ground vibration.
Concluding remarks
We performed spectral analyses of volcanic tremor and LF earthquake activity recorded
during the January-February 2005 eruption of Mt. Veniaminof, Alaska. Tremor had durations
of minutes to hours; we observed energy in the band 0.5-5.0 Hz, and spectra characterized by
sharp peaks irregularly or regularly spaced. Sustained non-harmonic tremor was interpreted as
the result of coalescence of large gas bubbles from a layer of smaller bubbles through the
surface of a magma column. We have suggested that the resonance of a gas-filled resonating
cavity may account for the occurrence and spectral features of harmonic tremor at Mt.
Veniaminof. Within this framework, the variable acoustic properties at the source or changes
in the effective length of the resonator, are possible causes for the observed time varying
characteristics of tremor such as spectral gliding.
Although the quality of present data at Mt Veniaminof is intrinsically limited by
instrumentation, there are several avenues for future research. The installation of new
instruments, such as broadband seismic sensors closer to the acive vent, may help to further
understand the source mechanisms of seismic signals and their propagation. Neverthless, the
19
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
January-February 2005 eruption of Mt. Veniaminof remains the best studied eruption of the
volcano to date, and provides a benchmark for comparing future and past eruptive activity.
Acknowledgements
The authors are grateful to the staff of the Alaska Volcano Observatory for their efforts
during the course of the eruption. We also thank two anonymous reviewers for insightful
comments that largely improved the manuscript. This work was supported by the Alaska
Volcano Observatory and the U.S.Geological Survey, as a part of their Volcano Hazards
Program, and by additional funds from the State of Alaska.
20
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
References
Aki K, Fehler M, and Das S (1977) Source mechanism of volcanic tremor: fluid driven cracks
models and their application to the 1963 Kilauea eruption. J Volcanol Geotherm Res 2: 259-
287
Aki K, Koyanagi (1981) Deep volcanic tremor and magma ascent mechanism under Kilauea,
Hawaii. J Geophys Res 86 B8:148-227
Bath M (1974) Spectral analysis in geophysiscs. Elsevier, Amsterdam.
Benoit J, McNutt SR (1997) New constraints on source processes of volcanic tremor at Arenal
volcano, Costa Rica, using broadband seismic data. Geophys Res Lett 24:449-452
Chouet B (1987) Resonance of a fluid-driven crack: radiation properties and implications for
the source of long-period events and harmonic tremor. J Geophys Res 93 B5:4375-4400
Crosson RS, and Bame DA (1985) A spherical source model for low frequency volcanic
earthquakes. J Geophys Res 90 B12:10237-10248
De Angelis S (2006), Analyses of unusual long-period earthquakes with extended coda
recorded at Katmai National Park, Alaska, USA. Geophys Res Lett 33 (in press)
Detterman RL, Miller TP, Yount ME, Wilson FH, (1981a) Geologic map of the Chignik and
Sutwik Island quadrangles, Alaska. Map I-1229, USGS
Detterman, RL, Miller, TP, Yount, ME, Wilson, FH, (1981b) Quaternary geologic map of the
Chignik and Sutwik Island quadrangles, Alaska Map I-1292, USGS
Fehler M (1983) Volcanic tremor at Mount St. Helens volcano. J Geophys Res 88:3476-3484
21
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
Fujita E, Ida Y, and Oikawa J (1995) Eigen oscillation of a fluid sphere and source mechanism
of harmonic volcanic tremor. J Volcanol Geotherm Res 69:365-378
Garces MA (1997) On the volcanic waveguide. J Geophys Res 102: 22547-22564
Garces MA, and McNutt SR (1997), Theory of the airborne sound generated in a resonant
magma conduit. J Volcanol Geotherm Res 78:155-178
Ginzburg IP (1963) Applied fluid dynamics. National Science Foundation, Washington DC.
Gordeev E (1993) Modelling of volcanic tremor as explosive point source in a single layered
elastic half-space. J Geophys Res 98: 19687-19703
Hagerty MT, Schwartz SY, Garces MA, Protti M (2000) Analysis of seismic and acoustic
observations at Arenal volcano, Costa Rica. J Volcanol Geotherm Res 101:27-65
Hellweg M (2000) Physical models for the source of Lascar’s harmonic tremor. J Volcanol
Geotherm Res 101:183-198
Kumagai H, Chouet B (2000) Acoustic properties of a crack containing magmatic or
hydrothermal fluids. J Geophys Res 25493-25512
Julian BR (1994) Volcanic tremor: nonlinear excitation by fluid flow. J Geophys Res
99:11859-11877
Julian BR (2000) Period doubling and other non-linear phenomena in volcanic earthquakes
and tremor. J Volcanol Geotherm Res 101:19-26
Miller TM, McGimsey RG, Richter DH, Riehle JR, Nye CJ, Yount ME, Dumoulin JA (1998)
Catalog of the historically active volcanoes of Alaska. 98-582, USGS
McNutt SR (1986) Observations and analysis of B-type earthquakes, explosions and volcanic
tremor at Pavlof volcano , Alaska. Bull Seismol Soc Am 76:153-175
McNutt SR (1994) Volcanic tremor from around the world: 1992 update. Acta Volcanologica
5:197-200
22
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
McNutt SR (2002) Volcano seismology and monitoring for eruptions, International handbook
of earthquake and engineering seismology, Volume 81 A, International Assoc. Seismol. &
Phys. Earth’s Interior, Committee on education, 383-406
Mori J, Patia H, McKee C, Lowenstein P, De Saint Ours P, Talai B (1989) Seismicity
associated with eruptive activity at Langila Volcano, Papua New Guinea. J Volcanol
Geotherm Res 38:243-255
Neuberg J, Luckett R, Baptie B, Olsen K (2000) Models of tremor and low frequency
earthquake swarms on Montserrat. J Volcanol Geotherm Res 101:83-104
Ripepe M, Gordeev E (1999) Gas bubble dynamics model for shallow volcanic tremor at
Stromboli. J Geophys Res 104: 10639-10654
Sánchez JJ, McNutt SR, Kumagai H, Gil-Cruz F, Londoño JM (2005) The complex
frequencies of long-period Seismic events at Mt. Veniaminof, Alaska. J Volcanol Geotherm
Res (submitted)
Sánchez JJ (2005) A velocity model for Mt. Veniaminof, Alaska: volcano-tectonic seismicity
and implications for an active magmatic system. J Volcanol Geotherm Res (submitted)
Sánchez JJ (2005) Volcano seismology from around the world: case studies from Pinatubo
(Philippines), Galeras (Colombia), Mount Wrangell and Mount Veniaminof (Alaska ). PhD
thesis, University of Alaska Fairbanks
Sassa K (1936) Volcanic microtremors and eruption earthquakes. Mem. Coll. Sci. Univ.
Kyoto Ser A 18:255-293
Schlindwein V, Wasserman J, Scherbabum F (1995) Spectral analysis of harmonic tremor
signals at Mt Semeru volcano, Indonesia. Geophys Res Lett 22:1685-1688
Simkin T, Siebert L (1994) Volcanoes of the world (2nd edition). Geoscience Press, Tucson
23
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
Thompson G, McNutt, SR, Tytgat G (2002) Three distinct regimes of volcanic tremor
associated with the eruption of Shishaldin volcano, Alaska 1999. Bull Volcanol 64:535-547
Tritton DJ (1988) Physical fluid dynamics. Clarendon press, Oxford
24
Date Color Code (*) Seismic observations Visual observations
4 Jan Upgraded to Yellow LF earthquakes; Non-harmonic tremor Ash explosions
7 Jan Yellow LF earthquakes; Non-harmonic tremor
Ash plumes (up to 3500 ft a.s.l.)
8-10 Jan Upgraded to Orange (Jan 10)
Seismic activity increases; LF earthquakes
Ash plumes
10 Jan -3 Feb Orange Continuous non harmonic tremor; Tremor bursts; LF earthquakes
Ash plumes
4 Feb Orange Non harmonic tremor; LF earthquakes
Incandescence from the intra-caldera cone (Feb 04, 05:32 UT)
5-16 Feb Orange Continuous non harmonic tremor; Tremor bursts
Persistent ash emissions
17 Feb Orange Tremor bursts; Harmonic tremor
Cloudy weather prevent the view of the intra-caldera cone
18 Feb Orange Continuous harmonic tremor Cloudy
19 Feb Orange Continuous tremor ends at 02:35 AST
Cloudy
20 Feb Orange No seismic activity until 11:00 AST when harmonic tremor picks up again
Cloudy
21 Feb Orange LF earthquakes; Tremor bursts; Harmonic tremor. Tremor abruptly ends at 14:00 AST
Cloudy
25 Feb Downgraded to Yellow
No relevant seismic activity Clear weather No activity observed
4 Mar Downgraded to Green No relevant seismic activity Clear weather No activity observed
(*) Level-of-concern color code definitions. To concisely describe the level of concern about possible eruptive activity at volcanoes, the Alaska Volcano Observatory has developed a color-code classification system (http://www.avo.alaska.edu/activity.php). Green: No eruption anticipated. Volcano is in a quiet, "dormant" state. Yellow: An eruption is possible in the next few weeks and may occurr with little or no additional warning. Small earthquakes detected locally and (or) increased levels of volcanic gas emissions. Orange: Explosive eruption is possible within a few days and may occurr with little or no warning. Ash plumes not expected to 25000 feet a.s.l. Increased numbers of local earthquakes. Extrusion of a lava dome or lava flows (non explosive eruptions) may be occurring. Red: Major explosive eruption expected within 24 hours. Large ash plume(s) expected to reach at least 25000 feet a.s.l. Strong earthquake activity detected even at distant monitoring stations. Explosive eruption may be in progress.
567
568
569
Table 1 Summary of seismic and visual observations during the January-February 2005 eruption of
Mt. Veniaminof, Alaska.
25
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
26
Figure 1. Seismograph stations (white squares) at Mt. Veniaminof on a shaded relief image (courtesy
NASA/JPL). The active intra-caldera cone is marked by a black triangle. In the bottom right corner,
an index map of Alaska shows the location of Mt. Veniaminof on the Alaska Peninsula.
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
Figure 2 Images from the January-February 2005 eruption at Mt. Veniaminof, Alaska. (a) Mt.
Veniaminof intra-caldera cone looking N-NW. Ash cloud drifting to NE. Photo taken at ~13,000 ft
from a Piper Navajo aircraft (Security Aviation) during an observational overflight (Photo taken on
January 11, 2005. Image Creator: K. L. Wallace, Image courtesy U.S. Geological Survey); (b) Mt.
Veniaminof intracaldera cinder cone. Ash plume drifting to NE. Photo source same as (a); (c),(d) and
(e) Ash plumes at Mt. Veniaminof. Images taken on January 09, 2005 from the AVO webcam located
in the town of Perryville (35 km SE of the active vent); c) 16:36:44 UT, d) 16:41:44 UT, e) 16:46:43
UT.
Figure 3. Seismograms (left) and amplitude spectra (right) of non-harmonic volcanic tremor recorded
during January 2005 at station VNSS. Average spectra (a), (b) and (c), were calculated by stacking
the Fourier spectra of six 10-second adjacent windows of data.
Figure 4. Seismogram (top) and spectrum (bottom) of a low-frequency event recorded on February
04, 2005, 06:06:15 UT, at station VNSS, during a period of Strombolian activity. Inset at lower right
is a stack of three 10-s windows of data.
Figure 5. Seismograms and velocity spectrograms of six hours of harmonic tremor, one hour per
panel, recorded at station VNSS on February 18, 2005. Spectrograms were generated moving a 1024
sample sliding window over the entire waveform and calculating the periodogram for 512 sample
overlapping positions of the window.
27
Figure 6. Seismograms (left) and amplitude spectra (right) of harmonic tremor recorded on February
18-19, 2005 at station VNSS. Average spectra (a), (b) and (c), were calculated by stacking the Fourier
spectra of six 10-second adjacent windows of data. Spectra are characterized by narrow and regularly
spaced peaks, which are integer multiples of 1.1 Hz.
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
Figure 7. Seismogram (top) and amplitude spectrum (bottom) of one minute of harmonic tremor
recorded at station VNSS on February 18, 2005. Five harmonics are clearly visible; inter-peak and
origin-to-first peak spacing is 0.6 Hz. Two additional peaks may be visible at 3.6 and 4.2 Hertz,
although their significance is lower.
Figure 8. One minute of harmonic tremor recorded on February 18, 2005 at four different stations
across the Mt. Veniaminof seismic network, and its power spectral density (PSD). PSD represents the
power content of a signal in an infinitesimal frequency band. Peaks at 1.1 and 2.2 Hz are well visible
at all stations; peaks at higher frequencies are attenuated at stations more distant from the active vent.
Figure 9. Seismogram and velocity spectrogram of harmonic tremor recorded at station VNSS on
February 18, 2005 showing gliding . t1, t2, t3 refer to time intervals discussed in the text.
Figure 10. Cartoon showing the evolution of activity at Mt. Veniaminof between late 2004 and
March 2005: a) variable conduit configuration between late 2004 and March 2005; b) free and c)
forced bubble coalescence; d) bubble bursting through the surface of the magma column; d) model of
resonant gas chamber. Representative waveforms associated with the different processes are shown
(non-harmonic tremor, low frequency event, harmonic tremor from the left to the right).
28
642
643
644
645
646
647
648
649
650 651 652 653 654 655 656 657 658
Figure 1. Seismograph stations (white squares) at Mt. Veniaminof on a shaded relief image
(courtesy NASA/JPL). The active intra-caldera cone is marked by a black triangle. In the
bottom right corner, an index map of Alaska shows the location of Mt. Veniaminof on the
Alaska Peninsula.
29
659 660 661
662
663
664
665
666
667
668
669 670 671 672 673 674 675 676
Figure 2 Images from the January-February 2005 eruption at Mt. Veniaminof, Alaska. (a) Mt.
Veniaminof intra-caldera cone looking N-NW. Ash cloud drifting to NE. Photo taken at
~13,000 ft from a Piper Navajo aircraft (Security Aviation) during an observational overflight
(Photo taken on January 11, 2005. Image Creator: K. L. Wallace, Image courtesy U.S.
Geological Survey); (b) Mt. Veniaminof intracaldera cinder cone. Ash plume drifting to NE.
Photo source same as (a); (c),(d) and (e) Ash plumes at Mt. Veniaminof. Images taken on
January 09, 2005 from the AVO webcam located in the town of Perryville (35 km SE of the
active vent); c) 16:36:44 UT, d) 16:41:44 UT, e) 16:46:43 UT.
30
677 678 679 680 681
682
683
684 685 686 687 688 689 690 691 692 693 694 695 696 697
Figure 3. Seismograms (left) and amplitude spectra (right) of non-harmonic volcanic tremor
recorded during January 2005 at station VNSS. Average spectra (a), (b) and (c), were
calculated by stacking the Fourier spectra of six 10-second adjacent windows of data.
31
698 699 700 701
702 703 704
705
706
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723
Figure 4. Seismogram (top) and spectrum (bottom) of a low-frequency event recorded on
February 04, 2005, 06:06:15 UT, at station VNSS, during a period of Strombolian activity.
Inset at lower right is a stack of three 10-s windows of data.
32
724
725 726 727 728
729
730
731
732
Figure 5. Seismograms and velocity spectrograms of six hours of harmonic tremor, one hour
per panel, recorded at station VNSS on February 18, 2005. Spectrograms were generated
moving a 1024 sample sliding window over the entire waveform and calculating the
periodogram for 512 sample overlapping positions of the window.
33
733
734
735
736
737
738 739 740
741
742
743
744 745 746 747 748 749 750 751
Figure 6. Seismograms (left) and amplitude spectra (right) of harmonic tremor recorded on
February 18-19, 2005 at station VNSS. Average spectra (a), (b) and (c), were calculated by
stacking the Fourier spectra of six 10-second adjacent windows of data. Spectra are
characterized by narrow and regularly spaced peaks, which are integer multiples of 1.1 Hz.
34
752 753 754 755 756 757 758 759 760 761 762 763 764
765 766 767 768
769
770
771
772 773 774 775 776 777 778 779 780 781
Figure 7. Seismogram (top) and amplitude spectrum (bottom) of one minute of harmonic
tremor recorded at station VNSS on February 18, 2005. Five harmonics are clearly visible;
inter-peak and origin-to-first peak spacing is 0.6 Hz. Two additional peaks may be visible at
3.6 and 4.2 Hertz, although their significance is lower.
35
782 783 784 785 786 787 788 789 790 791 792 793 794 795
796 797 798
799
800
801
802
803 804 805 806 807
Figure 8. One minute of harmonic tremor recorded on February 18, 2005 at four different
stations across the Mt. Veniaminof seismic network, and its power spectral density (PSD).
PSD represents the power content of a signal in an infinitesimal frequency band. Peaks at 1.1
and 2.2 Hz are well visible at all stations; peaks at higher frequencies are attenuated at stations
more distant from the active vent.
36
808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825
826 827 828
829
830 831 832 833 834 835 836 837 838 839 840
Figure 9. Seismogram and velocity spectrogram of harmonic tremor recorded at station VNSS
on February 18, 2005 showing gliding . t1, t2, t3 refer to time intervals discussed in the text.
37
841 842 843 844
845 846 847
848
849
850
851
852
853 854 855 856
Figure 10. Cartoon showing activity at Mt. Veniaminof between late 2004 and March 2005
and : a) variable conduit configuration between late 2004 and March 2005; b) free and c)
forced bubble coalescence; d) bubble bursting through the surface of the magma column; d)
model of resonant gas chamber. Representative waveforms associated with the different
processes are shown (non-harmonic tremor, low frequency event, harmonic tremor from the
left to the right).
38