4-2. Researches to Predict Volcano Eruptions
Intense, modern monitoring of volcanoes has
brought a considerable advance in the prediction of volcanic eruptions,
allowing some precursors to be identified prior to most eruptions. However,
accurate prediction of eruptions still cannot be achieved scientifically. The
following researches are being carried out to consolidate the solid foundation
of eruption prediction, based on better understanding of volcanic processes and
magma supply systems under volcanoes.
4-2-1. Joint Experiment on Volcano Structure
and Magma Supply System
Since 1994, joint experiments have been
conducted at several volcanoes in Japan to reveal the structure and magma
supply system by related national universities and scientists under the
National Project for Prediction of Volcanic Eruptions. Anomalous zones were
found beneath several volcanoes by electromagnetic surveys and seismic
exploration experiments (see, 6-12 Joint experiment on volcano structure). In
addition to the exploration experiments using controlled sources, we have also
developed new tomography methods using earthquakes. At Izu-Oshima volcano, we could
detect magma bodies at depths of 5 and 8-10km beneath the caldera with a new
technique of seismic waveform tomography, and also found a shallow high
velocity zone across the island in the NW-SE direction (a fossil of repeated
dike intrusion in the past) based on a cooperative inversion method using
travel-time and gravity data (see, 6-8 Volcano Research Center). At Fuji volcano,
we have started a joint experiment by constructing a dense seismic network of
100 stations to conduct basic researches on the origin of low-frequency
earthquakes and the volcano structure (see, 4-4 Fuji).
Fig.1 A
high velocity zone at a shallow depth across the Izu-Oshima volcano revealed by
a velocity-density cooperative inversion method.
4-2-2. Observations with ARGOS System
We have developed the telemetry system using ARGOS
satellite to obtain safely precise data of geomagnetic fields and fumaroles
temperature variations from the area closest to active craters. We deployed the
ARGOS geomagnetic observation system at the southwestern slope of Miyakejima
volcano just after the eruption on August 29, 2000, to detect geomagnetic total
force variations caused by temperature changes beneath the summit crater (Fig.2).
After July 2001, we detected a remarkable increase in total force intensity
corresponding to a decrease in the SO2 gas flux from the summit
crater (see, 4-6 Miyakejima eruption).
Fig.2 ARGOS geomagnetic observation system installed at
Miyakejima volcano.
4-2-3. Scientific Drilling at Volcanoes
We have conducted scientific drillings
at Izu-Oshima, Unzen and Fuji volcanoes, for the integrated studies of eruption
history, magma-plumbing system, subsurface structure and also high-quality
observation using boreholes. At Izu-Oshima volcano, we drilled a 1km-deep hole
within the summit caldera to detect small changes associated with the ascent of
magma and volcanic gas. Since 1999, we have started a 3D-array observation by
installing seismometers, a hydrophone and an electrical conductivity-temperature
meter in it (Fig.3). Borehole geophysical measurements and analysis of rock
samples yielded a new insight into the structure and origin of the caldera and the
eruption history of the volcano. At Unzen volcano, geological and petrologic
analysis of rock samples collected by drilling around the volcano yielded a new
insight into its eruption history and magma-plumbing system. In 2003, we started
the drilling into the summit conduit to obtain important information on
degassing processes during the magma ascent in the conduit (see, 4-5 Unzen). At
Fuji volcano, we drilled two 100m-deep holes and a trench at the NE and eastern
volcano slope, respectively. We also installed broadband seismometers in the
boreholes. In 2003, we will drill additional holes to install a broadband
seismometer, a tilt meter, and a 3-components strain meter (see, 4-4 Fuji).
Fig.3 Scientific drilling within the caldera of
Izu-Oshima volcano.
4-2-4. Researches on Mid- to Long-term
Predictions
We have conducted comparative studies
on the mid- to long-term predictions of volcanic eruptions. In Izu-Oshima,
which recently erupted every several tens of years, a continuous inflation of
the volcanic body, indicating the steady storing of magma at a depth of 6 to 9km,
has been detected by EDM and GPS observations. We also detected a continuous
inflation of Miyakejima volcano before the beginning of its activity on June
26, 2000 (see, 4-6 Miyakejima eruption).
Even without any historic eruption
records, a geological analysis of the volcano can sufficiently help us with
long-term predictions of eruptions (Fig.4). As the modes of eruption may
reflect magma compositions, analyzing products erupted in the earlier stage is also
important to know the mode of successive eruptions.
Fig.4
Cumulative volume-time diagram for eruption products at Fugen-dake in Unzen
volcano.
4-2-5. Enhancement of Cooperative Researches
on Eruption Prediction
A better cooperation among related
institutions and scientists is essential to promote researches on eruption
prediction, especially with the reorganization of national universities scheduled
in 2004. We will strengthen planning and coordinating functions to promote
joint experiments, education and quick response to volcanic eruptions in the
world, in cooperation with the Committee for Eruption Prediction Research
established in 2001 and visiting professors for prediction research newly
introduced in 2003 at ERI.