A. Arciniega-Ceballos | USGS Science

USGS2002

Long-period events and tremor at Popocatepetl volcano (1994–2000) and their broadband characteristics

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.

Bulletin of Volcanology↗

USGS2000

Temporal and spectral characteristics of seismicity observed at Popocatepetl volcano, central Mexico

Popocatepetl volcano entered an eruptive phase from December 21, 1994 to March 30, 1995, which was characterized by ash and fumarolic emissions. During this eruptive episode, the observed seismicity consisted of volcano-tectonic (VT) events, long-period (LP) events and sustained tremor. Before the initial eruption on December 21, VT seismicity exhibited no increase in number until a swarm of VT earthquakes was observed at 01:31 hours local time. Visual observations of the eruption occurred at dawn the next morning. LP activity increased from an average of 7 events a day in October 1994 to 22 events per day in December 1994. At the onset of the eruption, LP activity peaked at 49 events per day. LP activity declined until mid-January 1995 when no events were observed. Tremor was first observed about one day after the initial eruption and averaged 10 h per episode. By late February 1995, tremor episodes became more intermittent, lasting less than 5 min, and the number of LP events returned to pre-eruption levels (7 events per day). Using a spectral ratio technique, low-frequency oceanic microseismic noise with a predominant peak around 7 s was removed from the broadband seismic signal of tremor and LP events. Stacks of corrected tremor episodes and LP events show that both tremor and LP events contain similar frequency features with major peaks around 1.4 Hz. Frequency analyses of LP events and tremor suggest a shallow extended source with similar radiation pattern characteristics. The distribution of VT events (between 2.5 and 10 km) also points to a shallow source of the tremor and LP events located in the first 2500 m beneath the crater. Under the assumption that the frequency characteristics of the signals are representative of an oscillator we used a fluid-filled-crack model to infer the length of the resonator.

Journal of Volcanology and Geothermal Research↗