How do we prepare for infrequent huge volcanic...

Preview:

Citation preview

How do we prepare for infrequent huge volcanic eruptions?

Setsuya Nakada The University of Tokyo

2014/11/28 1

Preface and contents

Experiencing the 2011 Tohoku Large Earthquake, less-frequent huge volcanic eruption also became paid attention by the society. Especially, the impact from huge volcanic eruptions became the final key issue to re-run the Nuclear Power Plants in Japan. In this talk, I would like to share the research background and present situation for forecasting them. 1. Recent volcanic eruptions 2. Regularity: volcanic explosivity index vs. eruption frequency 3. Diverse volcanic hazards 4. Monitoring and forecasting volcanic eruptions 5. Caldera eruptions and precursory phenomena 6. Nuclear Power Plant issue 7. Conclusions and suggestions

2014/11/28 2

Volcanology

• Volcanology is the study of volcanoes, lava, magma and related geological, geophysical and geochemical phenomena. (--geology, physics, chemistry, astronomical science, philology, archaeology, environmental science, social science, etc.)

• International Association of Volcanology and the Chemistry of Earth’s Interior (IAVCEI): The primary international focus for: (1) research in volcanology, (2) efforts to mitigate volcanic disasters, and (3) research into closely related disciplines, such as igneous geochemistry and petrology, geochronology, volcanogenic mineral deposits, and the physics of the generation and ascent of magmas in the upper mantle and crust.

2014/11/28 3

Mount Ontake eruption in September 2014; small eruption (VEI=1) but many casualties

2014/11/28 4

Asahi Shimbun

No precursory!

2014/11/28 5

Courtesy by Asahi Shimbun

JMA

JMA

BL type

BH type

A type

Dai

ly n

umbe

rs

Tilt change

Seismic signal (up-down)

Sep 1

Oct 1

Oct 1

Sep 1 Oct 1

11:52 11:45

•Seismic swam 17 days before. •LF event continued. •Crater inflation 4 min before.

•Small signals, different modes among eruptions. •Difficult forecasting of small eruptions.

Nishinoshima (2013-14) An continuous moderate eruption but limited access

Taken from Asahi Shimbun aircraft on

Nov. 13, 2014

1,000 km South of Tokyo

2014/11/28 6

Nov. 24, 2013

Chaitén, Chile: May 2008 VEI=4-5

• No eruption record since 9,400 years ago, not monitored.

• Seismicity activated just one day before eruption at a “remote” station (300 km).

Buenos Aires

2014/11/28 7

www.dailymail.co.uk

NASA (Image of the day)

Recent large (not huge) eruptions

2014/11/28 8

AFP/Getty Images Puyehue-Cordón Caulle, erupted in June 2011 VEI=4

NASA Image of the day

Tasmania

2014/11/28 9

Sinabung (N Sumatra) Dec. 2013~ Magma volume ~0.1 km3

Sinabung, Sumatra Kelud, Java

Kelud (Java) Feb. 13, 2014 (VEI=4) Magma volume ~0.12 km3

Eruption diversity in term of discharge rate

When discharge rate is high, the eruption is explosive.

Explosive eruptions

Lava flow and dome eruptions

Disc

harg

e ra

te (

m3 /

s)

Eruption volume (DRE 106 m3)

106

104

102

100

10-2

10-3 10-1 101 103 105

Kelud 2014

Shinmoe-dake 2011

Shinmoe-dake 2011

Unzen 1991-95

Sinabung 2014

Nishinoshima 2013-14

Original diagram by Kozono et al. (2013)

High discharge rate (roughly, high rate of magma ascent) Magma’s degassing illness introduces explosive eruption.

Ash of explosive eruption

2014/11/28 10

Regularity: volcanic explosivity index vs. eruption frequency

2014/11/28 11

Magnitude of volcanic eruptions

2014/11/28 12

Huge

Moderate

Small

Large

Regularity of volcanic eruptions

1. Eruptions have not occurred randomly in the statistic point of view.

2. Regularity between the magnitude (VEI) and frequency can be seen in different regional scales.

3. Among Indonesia, Japan and Chile, Indonesia is most active.

Lengths of volcanic belts Chile: 4000 km Indonesia: 3000 km Japan: 2500 km

Small Moderate Large Huge Huge eruption: Caldera-forming eruption

Abnormally quiet in recent Japan

Very quiet in Japan during these centuries! Large eruptions are very possible near future in Japan. Annual probability for VEI=6 to 7 eruptions in Japan is =10-4. The latest VEI=7 eruption in Japan is 7,300 years ago.

The largest eruption for these 100 thousand years Toba caldera eruption, 74,000 years ago

(~ 5 cm isopach of fallen ash)

Ash falling area of Toba eruption

In 1991

In 1815

2014/11/28 15

Impact of the Kikai caldera eruption 7,300 years ago

Erupted magma: >85 km3

All eruptives: >170 km3

Volcanic ash falling area

Tsunami area? Pyroclastic flow area

Courtesy by F. Maeno

2014/11/28 16

VEI=7

Diverse volcanic hazards

2014/11/28 17

• Ash fall • Pyroclastic flow/surge • Lava flow • Lahar (mud flow) • Debris

avalanche/landslide/flank failure

• Volcanic tsunami

Varieties of volcanic eruptions/hazards

Tephra: fragmental material produced by a volcanic eruption

2014/11/28 18

Volcanic Ash Fall and pyroclastic flow

Pompei--AD79 tragedy

Pyroclastic surge

2014/11/28

19

[Story] Roman People could survived during ash falling. However, they were killed by pyroclastic flow (surge). [Lesson] First priority is for: the area of pyroclastic flows. People can survive in area the ash falling, if prepared. However, the effect to the infrastructure and environment are different.

Volcanic ash problems

2014/11/28 20

Flashover on glass insulator coated with 3 mm-thick wet volcanic ash, by 40 kV. Watson et al. (2011)

In April 2010, volcanic eruption (Eyjafjallajokull, VEI=4) in Iceland under the glacier produced very fine volcanic ash, which drifted over the northern Europe, resulting in airports closures.

Sulfur dioxide gas combines with water from the atmosphere and the volcanic cloud to produce tiny droplets of sulfuric acid - sulphate aerosols • Stratospheric aerosols have residence times of 2-3 years • Tropospheric aerosols have very short residence times •Ash has very short residence times, but effects are relatively unexplored

Atmospheric effects of volcanic ash

From McCormick et al., 1995

2014/11/28 21

The Atmospheric transparency after Pinatubo eruption (VEI=6) in 1991

• The aerosols absorbed incoming sunlight, the global temperature dropped by ~0.6 °C

• The effect continued at least four years.

http://earthobservatory.nasa.gov/IOTD/view.php?id=1510

2014/11/28 22

Monitoring and forecasting volcanic eruptions

The present level of volcanic eruption forecasting: Abnormal phenomena can be detected on the volcanoes equipped with the sufficient monitoring instrument. Based on them together with our empirical knowledge, we can say that an eruption is approaching . The scenario (magnitude, duration, mode, etc.) is difficult to say strictly.

2014/11/28 23

Seismicity

• Data from Kyushu University, guided in Nakada et al. (1999).

Usu volcano, Hokkaido

Erup

tion

Erup

tion

Felt earthquakes in 1910

Large earthquakes in 2000

Unzen volcano, Kyushu

2014/11/28 24

Deformation (GPS)

Mauna Loa, Hawaii

Shinmoe-dake (Kirishima), Kyushu VEI=3 Nakada et al. (2013)

Shinmoe-dake

2014/11/28 25

GPS baseline

Seismicity

Example of ground deformation observed in caldera areas with In-SAR

2014/11/28 26

Swelling observed at Uturuncu, Southern Central Andes during 1996-2000. Hickey et al. (2013)

Swelling observed at Yellowstone Caldera during 1996-2000

Pinatubo 1991

2014/11/28 27

1991 eruption (VEI=6) at Pinatubo, Philippines

Plume height

Volcanic gas (SO2)

PF event (column-collapse)

end PF event (dome-collapse)

end

Explosive

Magmatic

Magmatic

end end

Phreatic No collapse

Magma ascent

Yes/No EQ-type change, num. incr. w/ accelerate deformation

Yes/No Local deform. + EQ increase

Modified from 2011 ver.

es/No eform. + EQ

ncrease

Yes/No Juvenile in ash

Since Dec. 30, ‘13 Late Dec. ‘13

Mid-Nov. ‘13

-Dec. ‘13

Aug-Sep. ’10 Sep-Oct. ‘13

Edifice collapse, followed by the above scenarios

Sinabung volcano, Indonesia

Lava flow/dome

Eruption Scenarios • Event trees are beng introduced for forecasting the scenarios

of approaching or on-going eruptions recently. • This probabilistic consideration became important to avoid

misinterpretation.

2014/11/28 28

Caldera eruptions and precursory phenomena

When is the magma stored under the volcano, and how can we detect the abnormality (forecast the caldera eruption)?

2014/11/28 29

VEI 9 8 7 6 5 4 3

Caldera eruption

Repose time and magma accumulation rate

2014/11/28 30

F. Costa (2008)

Magma accumulation in the caldera volcanoes is faster than that in normal volcanoes.

Modeling of elemental diffusion and annealing in crystals Druitt et al (2012), Gualda et al. (2012) and so on.

Recent researches show that magma filled the chamber a hundred to thousands year before the climactic eruptions.

Magma accumulation time Examples in Santorini, Greece and Long Valley, USA

2014/11/28 31

Residence time deduced by crystal Chanbai-shan (9th Century, VEI=6)

Zou et al (2010) Lithos 119, 289-296.

U-Th disequiliburium age: ~ 9,000 years. As the crystal , zircon, may be recycled. the age determined is the maximum time. Accumulation rate = 107 to 6 m3/year

Pinatubo 1991

5 km3

2.5 km

10 km3 Krakatau 1883

8 km

~100 km3

鬼界 7300年前

20 km

~300 km3

Aso 90,000 years ago

20 km

1000 km3

Yellowstone 640,000 years ago 2800 km3

Toba 74,000 years ago

Thickness of magma stored in the crust before caldera eruptions (d)

d=0.8 km d=0.4 km

d=0.4 km

d=0.2 km

d=1 km d=1 km

2014/11/28 33

60 km 70 km

d

Caldera size Erupted magma represents the upper <1 km.

Could those amounts of magma be stored geologically suddenly? If so, we can monitor this!

Geological model

Ground deformation at Aira Caldera Magma accumulated under here already?

∆V/dt = ~107 m3/y Takes ~104 yrs to get ~100 km3 of magma in this rate.

2013.3.28 34

Nakao et al. (2013)

Volume change during 2003 to 2012

Apr. 2003 to Apr. 2009

Sakurajima

Precursory of Super eruptions in Indonesia and Philippines

No catastrophic deformation has not been noticed, but clear precursory phenomena had been observed several months before the climactic eruption. 2014/11/28 35

Mar Apr

Tambora 1815 (VEI=7)

Krakatau 1883 (VEI=6)

Collapse

Collapse

EQs started

EQs started

Plume height (km)

Lava dome

Extensive hydrothermal activity

Extensive hydrothermal

Explosion

EQs started

May Jun Jul

Aug Sep

Mar Apr

May Jun Jul

Pinatubo 1991 (VEI=6)

Day before the climax

days

Takada and Furukawa (2014) Iwanami Kagaku, 84, (1), 64-68

Plume height (km)

• Surface manifestation (small explosion, high seismicity and thermal activity) occurs a few months before the climax.

• Becomes more clear weeks before—maybe long enough for mass evacuation.

1yr 10 yr 100 yr 1 kyr 10 kyr

Precursory and repose times

1 month before

1 year before

1 day before

1 hour before

Prec

urso

ry p

erio

d

Passarelli and Brodsky (2012)

Usu Miyakejima

Unzen Shinmoe

Repose time 2014/11/28 36

Ontake Pinatubo

Tambora

Krakatau

Fuji

Power Plant siting issue

2014/11/28 37

Map of Nuclear Power Plants in Japan

32 units 1 unit 2 units

Kashiwazaki-Kariwa

Higashidori (Tohoku)

Onagawa

Fukushima-2

Fukushima-1

Tokai

Hamaoka

Ikata

Sendai Genkai

Shimane

Takahama

Ohi

Mihama

Tsuruga

Shiga

Tomari

A A

A

A

A

A: ABWR/APWR

A

Higashidori (TEPCO) A *

A Ohma

A

In operation 23 units Under construction 1 unit License applied 2 units

PWR BWR Total 55 units 2 unit 4 units

Tokyo

Before 2011

Aso Volcano

Caldera eruptions in Japan for about

100 kyears

Towada 10 kys

Aira ash

Toya ash

Toya caldera 100 kyrs

Shikotsu caldera 32 kyrs

Kutcharo caldera 110 kyrs

Aso ash

Aira ash Kikai ash

Kikai caldera 7 kyrs

Ata caldera 85 kyrs

Aira caldera 29 kyrs

Yellow: pyroclastic flow field

Aso caldera 90 ka

2014/11/28 39

Target volcanoes in Sendai NPP

2014/11/28 40

Aira caldera

Sendai NPP

Genkai NPP

Aso caldera

Ata caldera

Kikai caldera

Kakuto-Kobayashi caldera

Volcanoes requested in monitoring

Wikipedia

The Aira caldera was left as the most capable volcano after screening and evaluation.

Pyroclastic-flow deposits from the Aira Caldera (29 k years ago)

2014/11/28 41

Geology Simulation by Kyushu Electric Power Company

Sendai

Distribution of the Ito pyroclastic flow deposits Yokoyama (2000)

Issues related to re-running of the Sendai NPP

[background] • It seems that the site evaluation was not strict under the government

decision (rerunning). As a result, the final safety was brought to over-dependence on the volcanic monitoring ability.

• The present volcanology cannot forecast a caldera eruption with the leading time as long as a few years, necessary for transporting spent fuel.

[Solutions] • Setting severe thresholds, the operating company should shut down the

reactor when the thresholds are reached, and transport the spent fuel soon. • Since the spent fuel has already been stored in the NPP sites, however, the

volcanological monitoring and preparation for their transportation should be begun without the judgment of re-running.

2014/11/28 42

Natural disasters and researchers

• A natural disaster will happen, as long as the earth continues business and man lives. Preventing disasters is not the direct task of researchers.

• We have to recognize threats from nature from which we have received the blessing. There is no other way to minimize natural disasters than preparing for them upon understanding the society’s vulnerability.

• Natural Hazard Research is for understanding the processes, the structures, and results. It provides the wisdom to the society for preparing for the hazards.

• The knowledge should not be over-depended by the society but be utilized effectively. If not, nature’s retaliation is waiting for us in future.

2014/11/28 43

Conclusions and suggestions • The leading time to the climactic eruption, when forecasted, may be

weeks to months, which is long enough for people’s evacuation, if prepared sufficiently, but too short to start the preparation since then.

• Forecasting caldera eruption is the international issue. • Monitoring technique and modeling of caldera eruptions should be

improved not only by volcanologists but also scientists with various fields, including young scientists.

• Though the probability of huge volcanic eruptions is “VERY” low, it is not late to begin discussion on the situations which we would face.

• It is not too early to start discussion on forecasting and mass evacuation among scientists, economists, politicians and so on in the international level.

2014/11/28 44

Recommended