Transcript VOBPpresent_nakadax

Volcano Observatory Best Practice Workshop
- Near Term Eruption Forecasting
Erice, Sicily (IT), 11 - 15 September 2011
Unrest that led to eruption:
Unzen and Kirishima, Japan
Setsuya Nakada1 and Hiroshi Shimizu2
1. Volcano Research Center, Earthquake Research Institute, The
University of Tokyo
2. Institute of Seismology and Volcanology. Kyushu University
Eruption history at Mount Unzen
1663 - 1664: More than 30 people were killed by earthquakes and
debris flows.
1792: Failure of Mt. Mayuyama (old lava dome) generated tsunami.
About 15,000 people were killed.
1990 - 1995: Lava effusion continued for almost 4 years.
(2x108 m3 of dacite lava was extruded)
44 people were killed by pyroclastic flows.
About 2,500 houses were destroyed.
Mt. Fugen-dake
(Main Peak of
Unzen Volcano)
Mt. Mayuyama
Shimabara City
Precursor of eruption (seismicity)
Nov. ‘89
Dec. ‘89
Nov. 19, 1990
July ‘90
Oct. ‘90
Precursors of first eruptions
Earthquake
Volcanic
Tremor
Micro-pumice in phreatic eruption
products in February 1991
May 3, 1991
Precursors of
lava dome
emergence
These were reported to
the Coordinating
Committee for
Prediction
of Volcanic Eruptions
(CCVEP)
EDM
Ground tilt
Magnetic
total Force
May 20, 1991
GPS at Unzen
Nishi et al, (1999) JVGR v.89
Transmission of volcanic information in Japan
JMA only can issue official statements on
volcanic activity.
Japan Meteorological Agency (JMA)
Volcanic Information
Coordinating Committee for Prediction of
Volcanic Eruptions (CCPVE)
Monitoring
data
Local
government
National
Institutes
JMA
Observatories
of universities
Observation data/results are reported to
CCPVE which assesses the unrest.
Mass
media
Volc. Inform.
The public /
concerned inhabitants
Volcanic information is transmitted to
The public through local government/
mass media.
The Yomiuri
Pyroclastic flow event
•
•
Dome collapse started on May
24, 1991
Pyroclastic surges attacked
mass media and fire station
staffs. June 3, 1991
June 24, 1993
June 3, 1991
The prefectural governor asked the
Self Defense Force (SDF) for rescuing
casualties.
The operation in a limited area needed
real-time information on volcanic
activity.
Other civil protection agencies also
needed real-time information for quickly
respond to coming disaster.
The official information flow was too slow
during pyroclastic flow events.
Transmission of volcanic information at Unzen Volcano
Cooperation between
organizations concerned
made it possible to transmit
unofficial but useful volcanic
information: more quickly,
accurately and
understandably.
Japan Meteorological Agency
Official information
Coordinating Committee for
Prediction of Volcanic Eruptions
(CCPVE)
Local
governments
Unofficial information (real-time)
Mass
media
Shimabara Comments
Observatory
Self
Defense
Force
Advices
Police
Mass
media
Inhabitants
Local government
Cable TV
Unofficial information (real-time)
Members of SDF and police stationed at the observatory (SEVO),
watching seismograms and video monitors all day.
They transmitted the monitoring data directly to their headquarters.
The information was shared with the local government and cable-TV.
Volcanologist
Police
SDF, police, local governments and
inhabitants were able to be informed
immediately what was monitored.
SDF
Daily observation flight by SDF-helicopter
In addition, volcanologists shared the observation
data to SDF, police, local government and mass
media soon after helicopter flights every day.
Prof. Ohta
Local government official
SDF
Mass media
Mass media braodcasted
volcanologists’ comments soon
after the daily inspection flight.
SDF supported volcanologists in helicopter flights, Doppler radar observation,
and maintaining the observation system within the limited areas.
 Doppler radar was used to know
the travel distance of pyrolastic flows
IUGG (Melbourne) on July 4, 2011
Eruption at Shinmoedake (Kirishima) in 2011
Jan.26 pm
(Courtesy by Kazuo Shimousuki)
Location of Kirishima
Volcano Group
Kirishima
Pumice eruption in
Shinmoedake crater after
about 300 yrs silence.
In 1715-16, plinian
explosions with pyrolcastic
flows continued for two
years
Three sub-plinian explosions in Jan. 26 and 27, 2011
Jan 26 evening
Courtesy by Kazuo Shimousuki
Jan 27, 15:41 explosion
Lava accumulation in crater for Jan. 28-31, 2011
Explosion crater was sealed
with new lava.
Taken by Tetsuo Kobayashi on Jan 31, 2011
SAR images with a few
days interval were very
effective
Explosion crater was
covered completely
with new lava
|
Insufficient degassing
|
Highly possible strong
explosions
|
This observation data
were not reflected to
evacuation plan.
Entered into vulcanian stage
Courtesy by PASCO Co., Ltd.
TerraSAR-X
Strain change & magma volume
Vulcanian St.
Subplinian
explosions
Direction toward the source
Lava
accum.
stage
Normal direction
toward the source
Tephra
DRE,
x104 m3
Lava
accum.
DRE,
x104 m3
Deposit
730~
1100
1400
2100
~2500
Strain
1300
1400*
2600
Methods
Data of extensometer (Isa Observation
Station of DPRI, Kyoto Univ.)
Total DRE,
x104 m3
Eruption rates change
Plinian St.
Lava accum. St
Vulcanian St.
Typical
subplinian
Shinmoedake
2011
Intensity
~106 kg/s
~106 kg/s
Magnitude
~1011 kg
0.3-1.5x1010
kg
Column height
<20 km
7-9 km
Typical subplinian
explosion data from
Cioni et al. (2000)
Temporal change in seismicity in Kirishima Volcano
Yakiwara and others (2011)
Cumulative number
3000
2001/01/01 to 2011/06/30
3,401
2000
The rate
increased with
1000
time
0
4
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Number / Day
Earthquakes
3
The rate increase
in mid-2006 and
2
the end of 2009.
1
0
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Temporal changes of GPS baselines
Graphs after reducing the effects of vapor in
air and regional tectonic movement.
Geogr. Surv. Inst. (Data for 119th CCVEP)
Inflation slowly during
2006 to 2007,
accelerated after the end
of 2009
Juvenile in tephra
Pumice found in
tephra of Jan 19
(0.5 mm across)
Phreatic explosions began in 2008, and repeated in
2010.
10% of pumice were observed in the Jan 19, 2011
product, a week before sub-plinian explosions.
Unzen (1)
1. What kind of eruption forecasting assessment?
Start of eruption (Nov 1990)
1) Elevated seismicity and its migration
2) Clear attenuation of seismic waves passing under the summit
3) LP event that is the first time in monitoring at Unzen and increased.
Lava appearance (May 1991)
1) Swarm of high frequency B-type quakes beneath the crater
2) Rapid changes in EDM and tilt-meter and shallow demagnetization
3) Juvenile ash involved
2. How the forecasts have been achieved?
Couldn’t forecast exactly when steam explosion, but was expected.
Lava effusion was forecasted by CCVEP.
After lava effused (PF stage), rather qualitative assessment.
Unzen (2)
3. What kind or critical information was missed?
The manual to issue the alert was not prepared 20 years ago. Probably
better now…..?
4. How the scientific forecast has been used to take mitigation actions (the
decision-making chain)?
After lava effused, the official information flow was not useful due to
slowness.
Instead, communication of observatory scientists with the local
governments, mass media and army was effective.
5. The interaction between scientists, decision makers, and mass media.
Before lava effused, neither bad nor good.
After lava effused, on-site interaction among them went well and timely.
Volcanic warning introduced in 2007
Alert levels in Shinmoedake (Kirishima)
5: Evacuation
4: Prepare for evacuation
3: Limit approach to volcano (~2.5 km)
2: Limit approach to crater area (~1km)
1: Normal
Dates
Volcanic phenomenon Volcanic Alert issued
Aug. 22, 2008
Mar-Jul. 2010
2011
Jan. 19
Jan. 26-27
Jan. 29-31
Feb. 1-
Phreatic explosion
Phreatic explosions
Aug. 22-Oct. 29, 2008: level 2
Mar. 30-Apr. 16, Mar. 6, 2010: level 2
Magmatic eruption
Sub-plinian explosions Jan. 26: level 3 ~3km distance (bomb)..?
Lava accumulation
Jan. 31: level 3 ~3km (pyroclastic flow)..?
Vulcanian explosions
Feb. 1: level 3 ~4 km (bomb)
Mar. 22: level 3 ~3 km (bomb/pyr. flow)
Response was too slow
• A village decided evacuation by themselves in the night of Jan. 30, 2011.
• They lived within a few kilometers from the active crater, in the lowest side
without seeing the crater. Explosions that night were so noisy for them to
be very frightened.
• In addition, effusion of “lava dome” was observed two days before. The
word of “lava dome” made them to fall into a sort of panic, as they imaged
pyroclastic flow events at Unzen by it.
New lava dome in the crater floor (Jan. 28)
Shinomedake (Kirishima)
1. What kind of eruption forecasting assessment was made?
1) Inflation rate increased for a year
2) Rate of seismicity increased, though was not noticed correctly.
3) Precursory steam eruption for a few years
4) juvenile ash one week before the climax.
2. How the forecasts have been achieved
Forecast couldn’t be done correctly.
3. What kind or critical information was missed?
2) of 1 items. No one may have considered seriously.
4. How the scientific forecast has been used to take mitigation actions?
Personal scientific communication was useful in part.
5. The interaction between scientists, decision makers, and mass media.
Understanding and information issue were taken behind the phenomena.
Present condition of Shinmoedake
GPS
Daily cumulative time
of volcanic tremor
hrs
SO2