Jurgen Neuberg

Models of tremor and low-frequency earthquake swarms on Montserrat

Recent observations from Soufriere Hills volcano in Montserrat reveal a wide variety of low-frequency seismic signals. We discuss similarities and differences between hybrid earthquakes and long-period events, and their role in explosions and rockfall events. These events occur usually in swarms, and occasionally merge into tremor, an observation that can shed further light on the generation and composition of harmonic tremor. We use a 2D finite difference method to model major features of low-frequency seismic signatures and compare them with the observations. A depth-dependent velocity model for a fluid-filled conduit is introduced which accounts for the varying gas-content in the magma, and the impact on the seismic signals is discussed. We carefully analyse episodes of tremor that show shifting spectral lines and model those in terms of changes in the gas content of the magma as well as in terms of a time-dependent triggering mechanism of low-frequency resonances. In this way we explain the simultaneous occurrence of low-frequency events and tremor with a spectral content comprising integer harmonics.

Characteristics and causes of shallow seismicity in andesite volcanoes

Recent observations from Soufriere Hills volcano in Montserrat reveal a wide variety of seismic signals. We focus here on characteristics of long-period events, and their role in the pressurization of the volcanic plumbing system. These events can occur in swarms and merge into tremor, an observation that sheds further light on the generation and composition of harmonic tremor. A two-dimensional finite-difference method has been employed to model major features of low-frequency seismic signatures and compare them with the observations. A depth-dependent seismic velocity model for a fluid-filled conduit is used that accounts for the varying gas content in the magma. We analyse episodes of tremor that show shifting spectral lines and model those in terms of changes in excess pressure due to unloading and degassing events. Potential trigger mechanisms are suggested and discussed.

Results from the Broadband Seismic Network on Montserrat

A digital broadband seismic network has been installed around Soufriere Hills Volcano on Montserrat. While several distinctive types of seismic events with frequencies ranging from 0.5Hz to 30Hz could be identified, the emphasis is on two types of low-frequency events which indicate the involvement of a fluid phase in the source mechanism: the so-called long-period events and the hybrid events. The latter occur in swarms with distinct periodicities of 4 to 12 hours and preceed major dome collapses and explosions. The swarms correlate very well with the tilt observed at the flanks of the volcanic edifice and, hence, can be linked to the pressurization of the magmatic system. Occasionally separate hybrid events merge and form harmonic tremor, which sometimes has a shifting spectral content. This reveals temporary changes in the source parameters. Low-frequency seismic signals on Montserrat are considered to be key parameters for the monitoring of the internal dynamics of the volcano.

Modelling the time-dependent frequency content of low-frequency volcanic earthquakes

Abstract Low-frequency volcanic earthquakes and tremor have been observed on seismic networks at a number of volcanoes, including Soufriere Hills volcano on Montserrat. Single events have well known characteristics, including a long duration (several seconds) and harmonic spectral peaks (0.2–5 Hz). They are commonly observed in swarms, and can be highly repetitive both in waveforms and amplitude spectra. As the time delay between them decreases, they merge into tremor, often preceding critical volcanic events like dome collapses or explosions. Observed amplitude spectrograms of long-period volcanic earthquake swarms may display gliding lines which reflect a time dependence in the frequency content. Using a magma-filled dyke embedded in a solid homogeneous half-space as a simplified volcanic structure, we employ a 2D finite-difference method to compute the propagation of seismic waves in the conduit and its vicinity. We successfully replicate the seismic wave field of a single low-frequency event, as well as the occurrence of events in swarms, their highly repetitive characteristics, and the time dependence of their spectral content. We use our model to demonstrate that there are two modes of conduit resonance, leading to two types of interface waves which are recorded at the free surface as surface waves. We also demonstrate that reflections from the top and the bottom of a conduit act as secondary sources that are recorded at the surface as repetitive low-frequency events with similar waveforms. We further expand our modelling to account for gradients in physical properties across the magma–solid interface. We also expand it to account for time dependence of magma properties, which we implement by changing physical properties within the conduit during numerical computation of wave propagation. We use our expanded model to investigate the amplitude and time scales required for modelling gliding lines, and show that changes in magma properties, particularly changes in the bubble nucleation level, provide a plausible mechanism for the frequency variation in amplitude spectrograms.

What makes a volcano tick—A first explanation of deep multiple seismic sources in ascending magma

At many volcanoes, low-frequency earthquakes have often been associated with the state of a volcanic system and have been employed for eruption prediction. Several models attempt to explain the generation of such earthquakes, but fail to describe their clustering in tight spatial swarms and their highly repetitive nature. We present a new model that not only explains the generation of a single event, but also accounts for the swarm behavior and cyclic activity. By considering magma rupture as a source mechanism of seismic events, we demonstrate that a change in conduit geometry is the most plausible cause for their generation. Our model matches the observed spatial and temporal behavior of low-frequency seismicity and contributes to the understanding necessary to provide estimates of magma ascent rates.

Time dependent features in tremor spectra

Abstract Harmonic spectral peaks are observed in the tremor spectra of many different volcanoes, and in some cases these spectral lines have been seen to change with time. This has also been observed for the tremor at the Soufriere Hills volcano on Montserrat, West Indies, where the spectral lines are sometimes seen to glide apart before an explosion. We propose a model of repeated triggering of low-frequency earthquakes to explain these gliding lines using the relationship δt =1/ δν , where δt and δν are time and frequency spacing, respectively, and investigate factors which can affect the observation of these spectral peaks. Noise and amplitude variation are shown to have little effect on the spectral peaks; however the time gap between events must be nearly constant over several events. An error with a standard deviation of 2% or less is required for the spectral lines to be observed in the frequency range 0.5–10 Hz. We can reproduce the gliding spectral lines from a specific tremor episode preceding an explosion by changing δt from 1 to 0.31 s over a time period of 12 min. Using this relationship and an Automated Event Classification Analysis Program (AECAP), we can monitor δt over a long time period. The AECAP also extracts other seismic parameters such as energy, duration and spectral characteristics. An initial comparison between low-frequency seismic energy and cyclic tilt shows a correlation between the two, but this does not hold for later cycles.

Prototype PBO instrumentation of CALIPSO project captures world-record lava dome collapse on Montserrat Volcano

This article is an update on the status of an innovative new project designed to enhance generally our understanding of andesitic volcano eruption dynamics and, specifically the monitoring and scientific infrastructure at the active Soufriere Hills Volcano (SHV), Montserrat. The project has been designated as the Caribbean Andesite Lava Island Precision Seismo-geodetic Observatory known as CALIPSO. Its purpose is to investigate the dynamics of the entire SHV magmatic system using an integrated array of specialized instruments in four strategically located ∼200-m-deep boreholes in concert with several shallower holes and surface sites. The project is unique, as it represents the first, and only such borehole volcano-monitoring array deployed at an andesitic stratovolcano.

Unique and remarkable dilatometer measurements of pyroclastic flow–generated tsunamis

Pyroclastic flows entering the sea may cause tsunamis at coastal volcanoes worldwide, but geophysically monitored field occurrences are rare. We document the process of tsunami generation during a prolonged gigantic collapse of the Soufriere Hills volcano lava dome on Montserrat on 12–13 July 2003. Tsunamis were initiated by large-volume pyroclastic flows entering the ocean. We reconstruct the collapse from seismic records and report unique and remarkable borehole dilatometer observations, which recorded clearly the passage of wave packets at periods of 250–500 s over several hours. Strain signals are consistent in period and amplitude with water loading from passing tsunamis; each wave packet can be correlated with individual pyroclastic flow packages recorded by seismic data, proving that multiple tsunamis were initiated by pyroclastic flows. Any volcano within a few kilometers of water and capable of generating hot pyroclastic flows or cold debris flows with volumes greater than 5 × 10 6 m 3 may generate significant and possibly damaging tsunamis during future eruptions.

The relationship between degassing and rockfall signals at Soufrière Hills Volcano, Montserrat

Abstract One of the most common types of seismic event recorded during the eruption of Soufrière Hills Volcano from 1995 to 1999 is known as a rockfall signal because signals recorded when rockfalls were observed on the dome are of this type. Evidence is presented that two seismic sources contributed to these events. The action of falling debris on the dome generated seismicity between 2 Hz and 8 Hz, while many rockfall signals also have a marked spectral peak between 1 Hz and 2 Hz. Deployment of a pressure sensor near the volcano has shown that the 1-2 Hz energy was associated with degassing at the surface of the dome; however, the relative timing of gas escape and seismic signal showed that the first was not the direct source of the second. Resonance of the magma conduit linked to degassing at the surface is invoked as a probable source for the 1-2 Hz seismicity. The relative importance of the two seismic sources that contributed to rockfall signals is examined in the context of the behaviour of the volcano.

A model of the seismic wavefield in gas-charged magma: application to Soufrière Hills Volcano, Montserrat

Abstract Abstract: A stationary two-phase magma model is used to derive the relationship between gas mass and volume fractions, gas and magma bulk densities, temperature and pressure as a function of depth. In turn, these parameters are used to obtain the vertical seismic velocity profile in the gas-charged magma. A two-dimensional finite-difference model of a magma-filled conduit embedded in an elastic medium is then employed to generate the seismic wavefield in and around the conduit. The high impedance contrast between gas-rich magma and surrounding rock results in the seismic energy being efficiently trapped in the conduit; this leads to the generation of a long-lived resonance of tens of seconds commonly observed as low-frequency earthquakes and harmonic tremor. During a single seismic event, a variety of different seismic radiation patterns along the conduit is observed, leading to the occurrence of several distinct seismic phases in the synthetic seismograms. Observations from several volcanoes show peaked amplitude spectra with integer harmonic overtones that exhibit a time-dependent gliding. These features are successfully modelled by varying the excess pressure and, consequently, the gas volume fraction and the seismic velocity, representing sudden degassing events, such as the reduction of pressure by ash-venting, a dome collapse or a Vulcanian explosion, or, in turn, the pressurization of the conduit prior to such events.