Earth Sciences 240A - Lecture 24 - Volcanic Eruption Prediction

Prediction possible?

Activity Classification

Active

Repose

Dormant

Extinct

Deaths

Approx. 50 land-based eruptions/yr ; Only 1% of eruptions in last 100 years have caused any deaths (total: 70,000)!

8 May 1902: St. Pierre: 30,000 deaths: Pyroclastic flows

13 November 1985: Armero (Columbia): 25,000 deaths: lahars

Predictions

Short-Term

Long-Term

Primary participants of surveillance

USA; Russia; Japan; Iceland; Italy

Short-Term Prediction

Must be: Clear, precise; Short time; Danger zones defined; Type of danger expected

Search for Precursors
Seismic

Rising magma always makes earthquakes

1959-60 eruption

August: magma at 55 km; Gradually moved up

November: 1000 quakes/day; fissure opened

Short-term/long-term quakes

Hawaii: 51 seismic stations, automated

Ground deformation (tilt)

Tiltmeter: 1ppm sensitivity; Work well where eruptions localized

GPS: 3d images

Gas Evolution

SO2 alone not enough

SO2/CO2 progress

Summit lakes

Chemistry, pH, colour

Thermal Anomalies

Infrared detectors

Landsat images

Central Andes case:

Thought to be 16 ‘hot’ volcanoes; images showed 60

Earth Tides

Many small eruptions at Mount St. Helens since 1980 (building the dome)

‘Every one has been predicted to within 1 day’

Tracking quakes

Using deformation from Earth tides

From 1980-1987 17 eruptions took place; 14 predicted by Earth tides

Long-Term Prediction

Hazard map construction

Knowing plate boundaries, hot spot locations, geology

Danger assessment

Predicting an eruption is not the same as predicting a danger

E.g. Mt. Etna vs. Mt. Pinatubo

Mt. Nevado del Ruiz, Columbia (Armero City)

1595

1845

1985 (25,000 died)

Historical records

Great of frequent eruptions

Geological records

Most valuable; But what if no trace?

Potentially dangerous volcanoes in USA:

Mt. Baker, Mt. Ranier, Mt. St. Helens, Mt. Hood, Mt. Shasta, Lassen Peak

Are volcanic eruptions on the rise?

Next: Flood Basalts