s.eves@sstl.co.uk March 2008

Earthquake Prediction from SpaceStuart Eves – SSTL

with thanks to:- Mino and Friedemann Freund - NASA

Disaster Monitoring vs Disaster Mitigation

• SSTL has worked with nations around the globe to build the Disaster Monitoring Constellation (DMC)

• Ideally, though, SSTL and NASA would like to provide warning using space-based sensors before disasters strike

• For earthquakes, mitigation may now be possible

Seismic Signals

• Movements in the Earth’s crust can be detected by interferometric radar

• In general, however, it’s much easier to see the effects of earthquakes after they happen

ERS

Interferogram of the Bay of Naples area, showing volcanic changes in the topography around Vesuvius

E l e c t r o l y t e

+ + + ++ +

A n o d e C a t h o d e

A ARe ' e '

e ' e '

+…they act like a battery

Improved theoretical understanding

When rocks are compressed….

h•

h•

e’ e’

+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

p-type

n-type

Sourceh•

h•

h• h•

e’e’e’

e’e’

e’

e’

e’

e’

e’

h•h•h•

h•

h•

Man

tle

Con

tinen

tal C

rust

A rock battery

p-type

n-type

h•

h•S

IRemission

+ + + + + + + + + + + + + + + + + + + + +

IonosphereTEC

secondaryIR photons (8-10 µm)

primaryIR photons

930 cm-1

(10.7 µm)

h •

h •

h o t p e r o

x y l i n

k

Energy release at the surface

• Many different non-seismic signals may be produced• Which ones are real?• What do they tell us?

Non-seismic signals

GPSsatellite

Ionosonde

GPS receiver

Radio signals

Limb ScannerIRemission

VLF/ULFemission

VLF/ULF/ELFemission

Thermal Anomaly observed on 15 October 1999, 15 hrs before the Magnitude 7.1 Hector Mine, CA Earthquake

(courtesy Nevin Bryant, JPL)

Map with overlay graphicof surface traces (red and green)

of the fault.

20 Oct. 1998 Landsat image

(one yr before EQ).

old faults

old faults

new fault

new faultmaxstress

15 Oct. 1999 Landsat image

(15 hrs before EQ).Bright pixels are “hotter”.

Difference of 1999/1998Landsat images.

Bright pixels are “hotter”.

Infra-Red Detection

Map of IR anomaly Dec. 2003 versus Nov. 2003 for M=6.6 Bam earthquake (Iran) Dec. 26, 2003 (MODIS data).

Red star: Epicenter

Red line: Plate boundary

Brown lines: Major faults

(Ouzounov, Liu et al., 2007)

Iran

Iraq

SaudiArabia

Infra-Red Detection

Jan 20, 2001

Jan 21

Jan 22Jan 19

Jan 18

Jan 20

Pre-earthquake IR emission.MODIS data.

after Ouzounov & Freund Adv. Space Sci. 2004

M=7.6 Gujarat EQNW-India Jan. 26, 2001

Infra-Red Detection

p-type

n-type

h•

h•

h•

e’

S+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +

Closing the Battery Circuit

EMemission

IonosphereTEC

Brittle and weak

Brittle and strong

Ductile

Plastic

IonosphericPerturbation

Radio Frequency Detection

DEMETER

D

DataSummer 2006

Data combined from3561 earthquakes worldwide

M > 4.8

Evidence for EM emissionsin the 0-10 kHz range

shortly before and afterearthquakes

Demeter Satellite Data

“pearls” (Ionospheric

Tones)

Strong ELF Burst

Data compliments of UC Berkeley Seismological Lab

Parkfield M = 6 Earthquake Berkeley Data (0-10 Hz)

Day of quake - Sept 28, 2004

Radio Frequency Detection

Earthquake lights are:-Rare, short-lived, localWeeks or days before major earthquakesSometimes near epicenter, sometimes up to 100 km awaySometimes co-seismic Sudden luminous phenomenaLights coming out of the groundSometimes “curtains of light”Sometimes flickering flamesSometimes “ball lightning”

Obviously an electrical phenomenon.

Earthquake lights photographed by T. Kuribashi during 1966 Matsushiro earthquake swarm, Japan

Earthquake Lights

Possible Correlation with Lightning

• There are even some suggestions that earthquake-related electromagnetic activity might also affect the frequency of lightning strikes:-

— The lightning with which we’re familiar, that comes down from thunderclouds

— And just possibly the newly discovered forms of lightning, called elves and sprites, that shoot up from thunderclouds

Possible Earthquake Detection Satellite

• Satellite could be equipped with low frequency antennas to pick up low frequency radio events

• An array of three antennas on the satellite could measure the phase of incoming RF signals to provide accurate geolocation of the events

• Sensitive magnetometers and a limb sounder experiment to measure the total electron content of the ionosphere could also be of use

Uncooled Bolometer Array

• Sufficient to monitor small temperature changes on surface

•Low Power (<5W)

•100% duty cycle

• Compact

•Low cost

•Long lifetime

• The satellite could also be equipped with wide-area thermal sensors to detect the IR radiation

Constellation Issues

• In order to provide enough opportunities for observations in the days and hours before an earthquake, a constellation of satellites would be needed

• Initial estimates suggest that about 24 satellites, in relatively low (500 km altitude) orbits, might be required. This creates a requirement for an efficient propulsion system to counter drag

• Probable need for inter-satellite links or direct data downlinks to a network of ground stations to provide very rapid data delivery

• Possible need for on-board processing to provide near-real time warnings that can be passed through thin-route communication links

Conclusions

• Our understanding of the physical processes associated with earthquakes is improving

• There may be physical precursor signals that we can detect from space to provide warning of impending events

• Constellations of small satellites could collect these warning signals and save thousands of lives

Thank youFor further information contact :

Dr Stuart EvesSurrey Satellite Technology LimitedTycho House, Surrey Research Park,Guildford, Surrey, GU2 7YE, UKTel: +44 1483 803885, Fax: +44 1483 803804Email: s.eves@sstl.co.uk Web: www.sstl.co.uk

BACK-UP SLIDES

Laboratory Experiments

Laboratory Experiments

Laboratory Experiments

Laboratory Experiments

Earthquake Precursors

Earthquake Precursors

Earthquake Precursors

Laboratory Experiments