Sciency Thoughts: Predicting eruptions in monogenetic volcanic fields.

Seismic activity and fumerol (gas) emissions are well established as predictors of eruptions on stratovolcanoes (volcanic mountains which undergo repeated eruptions), but predicting eruptions in other kinds of volcanoes is more problematic. Monogenetic fields are areas of volcanic activity where sporadic eruptions occur at different locations, rather than at a single site. Most such sites only suffer very irregular eruptions, but the fields are often home to hot spring systems or other features which make them attractive to humans, leading to settlement in potentially hazardous areas, making finding a method for predicting eruptions in these areas a priority.

In a paper publishd in the journal Geology on 5 February 2016, Helena Albert of the Central Geophysical Observatory at the Spanish Geographic Institute, Fidel Costa of the Earth Observatory of Singapore and Asian School of the Environment at the Nanyang Technological University and Joan Martí of the Institute of Earth Sciences Jaume Almera, discuss a number of historic eruptions at a number of different monogentic fields, with a view to understanding the processes driving volcanic activity at these sites and the possibility of predicting future eruptions in such areas.

Albert et al. examined ten historic eruptions at five mongenetic fields; the 1704-05 Tenerife eruption (part of the Canary Islands Volcanic Field), the 1909 Tenerife eruption, the 1949 La Palma eruption (also in the Canaries), the 1971 La Palma eruption, the 2011 El Hierro eruption (again in the Canaries), the 1759 Michoacan eruption (in the Michoacan-Guanajuato region of Mexico), the 1943 Michoacan eruption, the 1943 Goropu Mountains eruption (part of the Owen Stanley Range of Papua New Guinea), the 1973 Heimaey Island eruption (Iceland) and the 1989 Higashi-Izu eruption (on the Izu Peninsula, Japan).

Emissions from the 2011 El Hierro eruption, which happened offshore to the south of the island. Guardia Civil.

Of these, only the 2011 El Hierro eruption was monitored with high quality modern seismic equipment, with three others having lower quality records available. However witness accounts of earthquakes provide insight into seismic activity in all of these fields prior to the onset of eruptive activity. All of the eruptions were proceded by periods of seismic activity, with the longest such recorded period preceding the 2011 El Hierro eruption, where tremmors were recorded 4-5 years ahead of the eruption, and the shortest being the 1973 Heimay Island eruption, where seismic activity was only noticed 2 days before the main eruption.

Lava flow from the 1973 Heimay Island eruption which entered the town of Vestmannaeyjar, destroying about half of the homes there and causing the population to evacuate to the Icelandic mainland. Will Perry/The Reykjavík Museum of Photography.

Albert et al. next looked at samples of erupted lava from all of the eruptions except the 1943 Goropu Mountains eruption and the 1989 Higashi-Izu eruption. In each case they found that the lavas were mineralogically mixed, that is to say included minerals thought to come from different magmatic intrusions. Magma extruded from deep within the Earth is thought to be more-or-less entirely liquid, however if it becomes trapped in chambers close to the surface (5-15 km underground) it cools foming a mush with crystals suspended in a liquid matrix. Different minerals are formed at different temperatures, and larger crystals form more slowly (implying the mush has remained cool enough for the crystals to form but hot enough for the matrix to remain liquid for a longer period of time), giving each magmatic intrusion its own distinctive makeup. Lavas with mixed compostitions contain crystals which cannot have formed in a single intrusion.

Ruins of the San Juan Parangaricutiro Church which was destroyed by the 1943 Michoacan eruption. Sparks Mexico/Wikipedia.

From this Albert et al. conclude that each of these volcanoc fields is fed by a an extensive shallow plumbing system with an extensive system of dykes at mid-crustal levels recieving intermittent intrusions of deep magma. While a full understanding of the processes behind this is beyond the scope of this study, the majority of these eruptions were preceded by a rise in seismic activity which began months to years before the main eruptive episodes, giving the potential for the prediction of such eruptions using dedicated seismic monitoring networks.

Possible plumbing system configuration and evolution of events that may occur below monogenetic volcanoes (schematic and not to scale). The depth of the subvolcanic system may vary from 5 to 15 km. The depth of the magma source is also variable but is at least 20 km. (A) Intrusion of magma approximately one or two years prior to the eruption, stalling of magma at 5–15 km due to the loss of buoyancy or freezing of dikes, and mixing processes registered by the crystals. Crustal assimilation occurs in some cases. Seismic activity is felt by the population in some cases. (B) Renewal of magma intrusion, progressive opening of the path between deep and shallow reservoirs, and mixing processes registered by the crystals. Crustal assimilation occurs in some cases. The seismic activity is commonly felt by the population. (C) Continued intrusion of mafic magma leads to easier transfer from deep to shallow reservoirs, and this allows the magma to finally erupt. Magma mixing (and crustal assimilation in some cases) are recorded by the crystals. Seismicity is felt by the population. Also shown the possibility that magma is directly transfer from the mantle to the surface. Albert et al. (2015).

Pages

Friday, 11 March 2016

Predicting eruptions in monogenetic volcanic fields.