Richard P. Hoblitt

Remarkable cyclic ground deformation monitored in real‐time on Montserrat, and its use in eruption forecasting

Telemetered high-resolution tiltmeters were installed in Montserrat in summer of 1995, in December 1996, and in May 1997. The 1995 installations, several km from the Soufriere Hills vent, were too distant to yield useful data. However, the 1996 and 1997 installations on the crater rim revealed 6–14 h inflation cycles caused by magma pressurization at shallow depths (< 0.6 km below the base of dome). The tilt data correlated with seismicity, explosions, and pyroclastic flow activity, and were used to forecast times of increased volcanic hazard to protect scientific field workers and the general public.

Cyclic eruptive behavior of silicic volcanoes

Silicic volcanism seems chaotic: the styles, magnitudes, and/or timing of successive eruptions may be without apparent pattern, or patterns may merge, fade, or abruptly change. However, geophysical monitoring of recent eruptions shows that some silicic volcanoes can exhibit cyclic eruptive behavior wherein periods of explosive activity or rapid extrusion alternate with periods of repose. The cycles are commonly observed in time-averaged amplitudes of eruption-related seismicity and also have been observed in ground-surface tilt data. When tilt and seismicity are both observed during oscillatory behavior, as at Soufriere Hills volcano on Montserrat, British West Indies, they correlate in time. Cycle periods range from hours to days, and cycle amplitudes and waveforms vary widely. Complex oscillatory behavior is also sometimes observed during high-pressure (tens of MPa) extrusion of industrial polymer melts. With this phenomenon as a guide, we construct a simple dynamic model for the oscillatory behavior of erupting volcanoes. We propose that cyclic eruptions result from Newtonian flow of compressible magma through the volcanic conduit combined with a stick-slip condition along the conduit wall, in analogy to the behavior of industrial polymers. If magma is forced into the conduit at a constant rate, pressure and flow rate rise. If the flow rate through the conduit exceeds a threshold value, the flow resistance abruptly drops as the magma slips along a shallow portion of the conduit wall. This reduces resistance to flow and causes the flow rate to jump to a higher value. If this enhanced flow rate exceeds the supply rate, both pressure and flow rate decline as the compressed magma in the conduit expands. Eventually slip ceases as the magma reattaches to the conduit wall at a flow rate less than the supply rate. Consequently pressure begins to increase, and the cycle begins again. This simple model reconciles a variety of disparate phenomena associated with cyclic silicic volcanism and provides a paradigm to interpret cyclic eruptive behavior.

Pyroclastic flows generated by gravitational instability of the 1996-97 lava dome of Soufriere Hills Volcano, Montserrat

Numerous pyroclastic flows were produced during 1996–97 by collapse of the growing andesitic lava dome at Soufriere Hills Volcano, Montserrat. Measured deposit volumes from these flows range from 0.2 to 9 × 106 m³. Flows range from discrete, single pulse events to sustained large scale dome collapse events. Flows entered the sea on the eastern and southern coasts, depositing large fans of material at the coast. Small runout distance (<1 km) flows had average flow front velocities in the order of 3–10 m/s while flow fronts of the larger runout distance flows (up to 6.5 km) advanced in the order of 15–30 m/s. Many flows were locally highly erosive. Field relations show that development of the fine grained ash cloud surge component was enhanced during the larger sustained events. Periods of elevated pyroclastic flow productivity and sustained dome collapse events are linked to pulses of high magma extrusion rates.

Mount St. Helens a decade after the 1980 eruptions: magmatic models, chemical cycles, and a revised hazards assessment

Available geophysical and geologic data provide a simplified model of the current magmatic plumbing system of Mount St. Helens (MSH). This model and new geochemical data are the basis for the revised hazards assessment presented here. The assessment is weighted by the style of eruptions and the chemistry of magmas erupted during the past 500 years, the interval for which the most detailed stratigraphic and geochemical data are available. This interval includes the Kalama (A. D. 1480–1770s?), Goat Rocks (A.D. 1800–1857), and current eruptive periods. In each of these periods, silica content decreased, then increased. The Kalama is a large amplitude chemical cycle (SiO2: 57%–67%), produced by mixing of arc dacite, which is depleted in high field-strength and incompatible elements, with enriched (OIB-like) basalt. The Goat Rocks and current cycles are of small amplitude (SiO2: 61%–64% and 62%–65%) and are related to the fluid dynamics of magma withdrawal from a zoned reservoir. The cyclic behavior is used to forecast future activity. The 1980–1986 chemical cycle, and consequently the current eruptive period, appears to be virtually complete. This inference is supported by the progressively decreasing volumes and volatile contents of magma erupted since 1980, both changes that suggest a decreasing potential for a major explosive eruption in the near future. However, recent changes in seismicity and a series of small gas-release explosions (beginning in late 1989 and accompanied by eruption of a minor fraction of relatively low-silica tephra on 6 January and 5 November 1990) suggest that the current eruptive period may continue to produce small explosions and that a small amount of magma may still be present within the conduit. The gas-release explosions occur without warning and pose a continuing hazard, especially in the crater area. An eruption as large or larger than that of 18 May 1980 (≈0.5 km3 dense-rock equivalent) probably will occur only if magma rises from an inferred deep (≥7 km), relative large (5–7 km3) reservoir. A conservative approach to hazard assessment is to assume that this deep magma is rich in volatiles and capable of erupting explosively to produce voluminous fall deposits and pyroclastic flows. Warning of such an eruption is expectable, however, because magma ascent would probably be accompanied by shallow seismicity that could be detected by the existing seismic-monitoring system. A future large-volume eruption (≥0.1 km3) is virtually certain; the eruptive history of the past 500 years indicates the probability of a large explosive eruption is at least 1% annually. Intervals between large eruptions at Mount St. Helens have varied widely; consequently, we cannot confidently forecast whether the next large eruption will be years decades, or farther in the future. However, we can forecast the types of hazards, and the areas that will be most affected by future large-volume eruptions, as well as hazards associated with the approaching end of the current eruptive period.

Bimodal Density Distribution of Cryptodome Dacite from the 1980 Eruption of Mount St. Helens, Washington

The explosion of a cryptodome at Mount St. Helens in 1980 produced two juvenile rock types that are derived from the same source magma. Their differences-color, texture and density-are due only to vesicularity differences. The vesicular gray dacite comprises bout 72% of the juvenile material; the black dacite comprises the other 28%. The density of juvenile dacite is bimodally distributed, with peaks at 1.6 g cm-3 (gray dacite) and 2.3 g cm-3 (black dacite). Water contents, deuterium abundances, and the relationship of petrographic structures to vapor-phase crystals indicate both rock types underwent pre-explosion subsurface vesiculation and degassing. The gray dacite underwent a second vesiculation event, probably during the 18 May explosion. In the subsurface, gases probably escaped through interconnected vesicles into the permeable volcanic edifice. We suggest that nonuniform degassing of an initially homogeneous magma produced volatile gradients in the cryptodome and that these gradients were responsible for the density bimodality. That is, water contents less than about 0.2–0.4 wt% produced vesicle growth rates that were slow in comparison to the pyroclast cooling rates; greater water contents produced vesicle growth rates that were fast in comparison to cooling rates. In this scheme, the dacite densities are bimodally distributed simply because, following decompression on 18 May 1980, one clast population vesiculated while the other did not. For clasts that did vesiculate, vesicle growth continued until it was arrested by fragmentation.