Source Processes Revealed at Two Guatemalan Volcanoes:

Insights from Multidisciplinary Observations of Harmonic Tremor and Numerical Modeling

Kyle A. Brill1(kabrill@mtu.edu) and Gregory P. Waite

1

Michigan Technological University

V33A-2845

PIRE Grant #0530109

I. Introduction Volcanic tremor (persistent ground motion occurring over a time range from minutes

to days) is typically associated with different types of material transport beneath a

volcano, and tremor that presents with multiple peaks (overtones) at integer multi-

ples, it is termed harmonic [Konstantinou and Schlindwein, 2003]. Harmonic tremor

can provide insight into the source processes that created them, but interpretation is

complicated by the complex volcanic environments.

Figure 1 shows station locations where we recorded instances of harmonic tremor at

Santiaguito and Fuego volcanoes in southern Guatemala, Central America.

Santiaguito

Santiaguito (2500 m) is a dacitic dome complex located in the 1902 eruption crater

of Santa María (3772 m). We recorded data at SG01 using a 3 component broadband

(30 second) seismometer for ~44 days of data from 1 January 7-April 2009 with data

gaps due to intermittent power losses. The data set was supplemented by infrasound

recordings, time-lapse images, thermal images, and video recordings.

Fuego

Fuego (3263 m) is a basaltic andesite composite volcano and marks the current erup-

tive center of the Fuego-Acatenango volcanic complex. We gathered broadband

seismic data at 10 stations around Fuego from 8 January-26 January 2009 along with

infrasound recordings, high-speed video recordings, and UV images.

II. Gliding Tremor Short-time window Fourier transform (STFT) plots

with an 8 second window and 80% overlap reveal

tremor with spectral peaks at integer multiples.

Both volcanoes demonstrated spectral gilding of the

harmonic frequencies. We quantify the gliding in

the signals using the fpeak function developed by

Geng Jun for Matlab [Jun 2003] to pick local mini-

ma and maxima within the spectrograms.

These rapid frequency fluctuations have direct im-

plications for source interpretation.

Most proposed models (thermo-acoustic resonators

[Busse et al. 2005], magma column wagging

[Jellinek and Bercovici, 2011], clarinet-organ pipes

[Lesage et al., 2006]) are not capable of producing

such dramatic frequency shifts over so short a time

without invoking geologically improbable geome-

tries or implausible situations. We rely on extra in-

formation from contemporaneously collected data

to aid in our interpretation.

Volcano f0 Range

(Hz)

Event

Duration

(seconds)

Santiaguito 0.3-0.7 Hz 38-1248

Fuego 0.5-2.5 Hz 30- > 3600

Volcano f0 Maximum

Rate of

Change

f0 Maximum

Total

Change

Santiaguito 0.03 Hz/sec 0.36 Hz/

Fuego 0.1 Hz/sec 0.75 Hz/

III. Tremor Source The best fit model to our data and observations at each volcano

is a repeating series of nonlinear excitations of fracture walls

as gas flows through the system after Julian [1994].

Differences in how cracks form and heal themselves account for

variability in activity.

Santiaguito Fuego

Magma rises and viscosity increases.

Strain rates exceed shear strength.

Magma fractures at the conduit margins.

Small cracks form fracture networks.

Gas escapes and excites fracture walls.

Pressures are released, fractures heal with added heat.

Undercooling of magma allows rapid crystallization.

Melt viscosities surpass normal range for basalt.

Cap forms and partially seals the system.

Gas escapes through partially plugged secondary vent.

Gas escapes and excites fracture walls.

Pressures are released, fractures heal with added heat.

IV. Particle Motions We see unexpected changes in the polarization of particle motions at both

Santiaguito and Fuego volcanoes. We use the frequencies identified by the

Find Peak Value algorithm and create 2-pole band pass filters on either

side of the chosen frequency. When the isolated frequencies are plotted,

the particle motions are typically elliptical and not easily attributed to a

single wave type. Many examples show 90° rotations in particle motions

from one overtone to the next. The relative amplitude contributions and

dominant directions also vary greatly within continuous events.

One possible explanation for these behaviors could be that the lower fre-

quency signals are within the near-field for those wavelengths, while the

higher frequencies are not. Signals between 0.3 and 2.0 Hz correspond to

wavelengths of 10 km or 1.5 km at a vp=3 km/s, our stations are mostly at

a 1-3 km distance. Lokmer and Bean [2010] highlight fine resolution fre-

quency contribution changes as well as polarized particle motions as the

near-field term decays. Exploring the details of where and how the near-

field term affects the wave field at about one wavelength distance could

shed more light on this phenomenon.

V. Conclusion At both Santiaguito and Fuego volcanoes, gliding harmonic tremor signals with highly varia-

ble particle motions point to sources which are unstable by themselves but formed by repeti-

tive processes capable of producing similar although not identical conditions. Pressurized

gasses passing through a changing crack network and exciting nonlinear responses from the

crack walls as proposed by Julian [1994] best explain these characteristics.

At Santiaguito, crack networks are created by shear fracture after Holland et al. [2011], while

at Fuego, cracks open through a secondary vent filled by cooled lava [Nadeau et al., 2011].

Acknowledgments Funding by US National Science Foundation PIRE Grant #0530109. Special thanks to the people at INSIVUMEH (Eddy Sanchez, Gustavo Chigna, Alvaro Rojas, Julio Cornejo, Amilcar Cardenas, Edgar Barrios) and

Peace Corps (Flavio Linares, Craig Badger) for the help and permission to do this work. Thanks also to Bill Rose, Seth Moran, Jeff Johnson, Jonathan Lees, John Gierke, John Lyons, Joshua Richardson, Rüdigger Esco-

bar Wolf, Emily Gochis, Jesse Silverman, Julie Herrik, Randall Fish, Hans Lechner, Rob Hegemann, Lara Kapalansik, Dulcinea Avouris, Justin Olson, Kelly McLean, Amie Ledgerwood.

References Busse, F. H., P. A. Monkewitz, and M. Hellweg (2005), Are harmonic tremors self-excited thermoacoustic oscillations?, Earth and Planetary Science Letters, 240(2), 302-306.

Holland, A. S. P., I. M. Watson, J. C. Phillips, L. Caricchi, and M. P. Dalton (2011), Degassing processes during lava dome growth: Insights from Santiaguito lava dome, Guatemala, J. Volcanol. Geotherm. Res., 202(1-

2), 153-166.

Jellinek, A. M., and D. Bercovici (2011), Seismic tremors and magma wagging during explosive volcanism, Nature, 470(7335), 522-525.

Konstantinou, K. I., and V. Schlindwein (2003), Nature, wavefield properties and source mechanism of volcanic tremor: a review, J. Volcanol. Geotherm. Res., 119(1-4), 161-187.

Lesage, P., M. M. Mora, G. E. Alvarado, J. Pacheco, and J.-P. Métaxian (2006), Complex behavior and source model of the tremor at Arenal volcano, Costa Rica, J. Volcanol. Geotherm. Res., 157(1-3), 49-59.

Lokmer, I., and C. J. Bean (2010), Properties of the near-field term and its effect on polarisation analysis and source locations of long-period (LP) and very-long-period (VLP) seismic events at volcanoes, J. Volcanol.

Geotherm. Res., 192(1-2), 35-47.

Lyons, J. J., and G. P. Waite (2011), Dynamics of explosive volcanism at Fuego volcano imaged with very long period seismicity, J. Geophys. Res.-Solid Earth, 116.

Nadeau, P. A., J. L. Palma, and G. P. Waite (2011), Linking volcanic tremor, degassing, and eruption dynamics via SO2 imaging, Geophysical research letters, 38.