Earthquake Early

Warning System Warning System

M. L. SHARMA, A. GAIROLA, KAMAL, RAVI

G. RATHORE, B. CHAMOLI, P. KUMAR, R.

Earthquake Early

Warning System Warning System

, KAMAL, RAVI JAKKA

, B. CHAMOLI, P. KUMAR, R. SACHDEVA

Seismic Hazard Vulnerability

Multi-dimensional

IntroductionRecent spurt in Seismicity; Seismic gaps; No earthquake prediction

Faster pace of development (unplanned), vulnerability increased

Seismic hazard and risk exercises emphasize need of disaster mitigation and management strategies

EEW can be used for alerting people few tens of seconds before arrival of damage causing waves.

Deterministic

Statistical

ProbabilisticClassical DSHAGeological evidencesGeophysical evidencesDeformation studiesPaleo seismic StudiesParametric StudiesPast data analysesClassical ApproachesConditional ProbabilitiesExtreme Value Statistics…………….

Multi-dimensional Physical, social, economic, environmental, institutional, human factors

Dynamicchanges over time

Scale-dependent expressed at different scales from human to household to community to country resolution

Vulnerability Risk

dimensional

earthquake prediction tools

Faster pace of development (unplanned), vulnerability increased

Seismic hazard and risk exercises emphasize need of disaster mitigation and management strategies

EEW can be used for alerting people few tens of seconds before arrival of damage

dimensional , social, economic,

environmental, institutional,

time

dependent at different scales

from human to household to community to country

Life Loss

Economic Loss

Seismic Hazard

Seismic Zone IV and V

Recent Past damaging earthquakes

Magnitude Return period

6 10 years

Date

04-04-1905

15-01-1934

15-08-1950

21-07-1956

11-12-1967

19-01-1975

06-08-1988

21-08-1988

20-10-1991

30-09-19936 10 years

7 50 years

8 200 years

Distances from Seismogenic sources

On the source

About 150 km from the sources

29-03-1999

26-01-2001

26-12-2004

08-10-2005

18-09-2011

25-04-2015

12-05-2015

1819-06-16

1845-06-19

1869-01-10

1897-06-12

Location Meg Intesity Deaths

1905 Kangra 7.8 Ms IX >20,000

1934 Nepal 8.0 Mw XI 6,000–10,700

1950 Assam, Tibet 8.6 Mw XI 1,500–3,300

1956 Gujarat 6.1 Ms IX 115

1967 Maharashtra 6.6 Mw VIII 177–180

1975 Himachal

Pradesh

6.8 Ms IX 47

1988 Myannmar, India 7.3 Mw VII 3

1988 Udayapur, Nepal 6.9 Mw VIII 709–1,450

1991 Uttarkashi, Uttara

khand

6.8 Mw IX 768–2,000

1993 Latur, Maharasht

ra

6.2 Mw VIII 9,748

ra

1999 Chamoli district-

Uttarakhand

6.8 Mw VIII ~103

2001 Gujarat 7.7 Mw X 13,805–20,023

2004 off northern

Sumatra

9.1–

9.3 Mw

IX 230,000–280,000

2005 Kashmir 7.6 Mw VIII 86,000–87,351

2011 Gangtok, Sikkim 6.9 Mw VII >111

2015 Nepal, India 7.8 Mw IX 8,964

2015 Nepal, India 7.3 Mw VIII 218

16 Gujarat 7.7–

8.2 Mw

XI >1,543

19 Runn of Kutch 6.3 Ms VIII Few

10 Assam, Cachar 7.4 Mw VII 2

12 Shillong, India 8.0 Mw X 1,542

Percentage of Houses with Different Level of

Risk in Major Towns near Central Himalayas

(BMPTC 2006)

Cities Benfitted by EEW and

Instrumented Cities (Census 2011)

Seismic hazard studies for damsSeismic hazard studies for damsSerial Project_Name

1 Jamrani- Nainital

2 Lakhwar Dehradun

3 Pithoragarh

4 Tehri

5 Dhauliganga- Pithoragarh

6 Srinagar-Garhwal

7 Kishu Dehradun

8 Koteshwar Teheri

9 Vyasi

10 Loharinag Pala

16 Kotlibhel II

17 Bhairon Ghati

18 Shrinagar

19 Rupsiabagar

20 Jamrani

21 Singoli-Bhatwari

22 Jakhol-Sankri

23 Alaknanda

24 Jelam Tamak

25 Malari Jelam

26 Bowala-Nand-10 Loharinag Pala

11 Tapovan Vishnugad

12 Koteshwar

13 Piplakoti

14 Kotlibhel IA

15 Kotlibhel IB

26 Bowala-Nand-

27 Devsari

28 Tiuni-Plasu

29 Lakhwar

Seismic Hazard, Vulnerability and

Risk values are quite

We are living very close to

seismogenic sources.

Need

seismogenic sources.

EEW system can save a lots of

lives and economic losses.

Real time earthquake information

collection from the sensors

Earthquake size estimation with

predictive result

EEW Components

predictive result

Warning dissemination to users

EEW Systems

Onsite Earthquake Early Warning System• Vibration sensors are placed at the site where warning has to be given.

• Warning based on single sensor or group of sensors deployed closely

• Basis: P Waves travels faster than damaging S Waves

• Chances of false and Missed warnings are too high

• Mostly used for railways, metros and industries

Onsite

Regional Earthquake Early Warning System• Vibration sensors are placed at the site where a large earthquake is expected to

occur and warning has to be given at a further distant area.

• Warning based on alerts from at least four to five sensors located at different

places

• Basis: Picking EQ near epicenter and use of EM waves to communicate information

• Chances of false and missed alarms are very little

• Can be used to save lives

Regional

Systems

Onsite Earthquake Early Warning SystemVibration sensors are placed at the site where warning has to be given.

Warning based on single sensor or group of sensors deployed closely

Basis: P Waves travels faster than damaging S Waves

Chances of false and Missed warnings are too high

Mostly used for railways, metros and industries – seismic switches ??

Regional Earthquake Early Warning SystemVibration sensors are placed at the site where a large earthquake is expected to

occur and warning has to be given at a further distant area.

Warning based on alerts from at least four to five sensors located at different

Picking EQ near epicenter and use of EM waves to communicate

Chances of false and missed alarms are very little

Can be used to save lives

EEW Systems: Lead TimeLe

ad

Tim

e

(Satriano et al. 2011)

Lea

d T

ime

Systems: Lead Time

Onsite

Regional

et al. 2011)

Onsite warning

issued after

arrival of P

wave

Lea

d T

ime

EEW Systems: Lead Time

(Satriano et al. 2011)

wave

Lea

d T

ime

Onsite

Systems: Lead Time

et al. 2011)

Onsite warning

issued after

arrival of P

wave

Lea

d T

ime

EEW Systems: Lead Time

(Satriano et al. 2011)

wave

After this point S

waves reached

after warning

Lea

d T

ime

Onsite

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

Onsite warning

issued after

arrival of P

wave

Lea

d T

ime

12

se

c

EEW Systems: Lead Time

(Satriano et al. 2011)

wave

After this point S

waves reached

after warning

Lea

d T

ime

12

se

c

Onsite

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

Lea

d T

ime

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime

Regional

Systems: Lead Time

et al. 2011)

Lea

d T

ime

Regional warning

issued after 10

secs of arrival of

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime secs of arrival of

earthquake

Regional

Systems: Lead Time

et al. 2011)

Lea

d T

ime

Regional warning

issued after arrival

of earthquake

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime

of earthquake

After this point S

waves reached

after warning

Regional

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

Lea

d T

ime

Regional warning

issued after arrival

of earthquake

25

se

c

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime

of earthquake

After this point S

waves reached

after warning

25

se

c

25

se

c

Regional

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

25

se

c

Lea

d T

ime

Regional warning

issued after arrival

of earthquake

25

se

c

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime

of earthquake

After this point S

waves reached

after warning

25

se

c

25

se

c

Regional

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

25

se

c

Lea

d T

ime

Regional warning

issued after arrival

of earthquake

25

se

c

EEW Systems: Lead Time

(Satriano et al. 2011)

Lea

d T

ime

of earthquake

After this point S

waves reached

after warning

25

se

c

25

se

c

Regional

Systems: Lead Time

et al. 2011)

After this point S

waves reached

after warning

25

se

c

EEW System

Japan Earthquake early warning system, USA shake maps and EEW,

Mexican Seismic alert system, P Alert Taiwan, Turkey, China, Italy, ………

EEW System

Japan Earthquake early warning system, USA shake maps and EEW,

Mexican Seismic alert system, P Alert Taiwan, Turkey, China, Italy, ………

Isoseismal map

~250 km

IX and VIIIIX and VIII

~600 km

X and IX

~200 km

VIII and VII

Intensity scale

Type of Type of Definitions Quant Percentage

VIIIVIII Most buildings of Type C suffer damage of GradeMost buildings of Type B suffer damage ofGrade 4. Many buildings of Type C suffer damage

IXIX Many buildings of Type C suffer damage of Grade

Many buildings of Type B show damage of Grade

Many buildings of Type A suffer damage of Grade

Type of Type of structustructu

Definitions

AA Buildings in fieldstone, ruralstructures, unburnt-brick houses,clay houses.

BB Ordinary brick buildings, buildingsof the large-block andprefabricated type, halftimbered structures, buildings innatural hewn stone.

CC Reinforced buildings, well built wooden structures

Quant Percentage

Single, few

About 5 %

Many About 50 %

Most About 75 %

Grade Definitions Descriptions

Grade 2, and a few of Grade 3.Grade 3, and most buildings of Type A suffer damage

damage of Grade 4.

Grade 3, and a few of Grade 4.

Grade 4, and a few of Grade 5.

Grade 5.

Grade Definitions Descriptions

G1 Slight damage

Fine cracks in plaster; fall of small pieces of plaster

G2 Moderate Small cracks in walls; fall of fairly damage large pieces of plaster, pantiles slip off; cracks in chimneys; parts of chimney fall down.

G3 Heavy damage

Large and deep cracks in walls; fall of chimneys.

G4 Destruction Gaps in walls; parts of buildings may collapse; separate parts of the building lose their cohesion; and inner walls collapse.

G5 Total damage

Total collapse of buildings.

Regional EEW for Northern India

Pilot project Initially funded by MOES

Deployed in 2015 with 84 stations streaming data in real

26 stations installed at blocks/tehsils/districts networked

Uttarakhand

58 stations have been installed inside BSNL towers and are connected using

Algorithm for real time processing of the data for earthquake early warning

Simulation for performance of the software successfully completed using previously recorded data of

Taiwan of similar instruments installed in similar conditions

BOLD STEP Sirens deployed in IITR campus

Regional EEW for Northern India

data in real time

blocks/tehsils/districts networked using SWAN (State Wide Area Network)

58 stations have been installed inside BSNL towers and are connected using VPNoBB

Algorithm for real time processing of the data for earthquake early warning were tested.

Simulation for performance of the software successfully completed using previously recorded data of

Taiwan of similar instruments installed in similar conditions.

Sirens deployed in IITR campus

Installation of sirens

Control Unit

Internet

Installation of sirens

Control Unit

Present Status

Taken over by Uttarakhand Government All sensors working successfully

Objectives of the new project

Maintenance of the 84 sensor network Deployment of 100 new sensorsMaintenance of the 84 sensor network

Expansion of the network towards Dharchula area

Deployment of 100 new sensors

Develop algorithms to issue warning to society

Installation of sirens at District Headquarters

Installation of sirens at State Emergency Operation Centres (SEOC), Dehradun and

Haldwani

Status

All sensors working successfully

Deployment of 100 new sensors

area

Deployment of 100 new sensors

Installation of sirens at State Emergency Operation Centres (SEOC), Dehradun and

EEW for Northern India

84 sensors deployed in Himalayas to cater to the need

Of Earthquake Early Warning system to cities of Northern India

EEW for Northern India

84 sensors deployed in Himalayas to cater to the need

Of Earthquake Early Warning system to cities of Northern India

Network used for EEW system

For the current set-up of EEW system for Northern India we

are using 2 different network for communication of real

time earthquake signals

26 sensors have been installed at the PoPs of SWAN

58 sensors have been installed at the BTS of BSNL using

VPNoBB.

For data from SWAN network we have installed a 2 MBPS For data from SWAN network we have installed a 2 MBPS

lease line connecting DHQ-Haridwar to IIT Roorkee

For data from sensors installed at BSNL BTS, we have

installed a 2 MBPS MPLS circuit that connects Dehradun to

IIT Roorkee.

Network used for EEW system

up of EEW system for Northern India we

are using 2 different network for communication of real

of SWAN

58 sensors have been installed at the BTS of BSNL using

For data from SWAN network we have installed a 2 MBPS For data from SWAN network we have installed a 2 MBPS

For data from sensors installed at BSNL BTS, we have

installed a 2 MBPS MPLS circuit that connects Dehradun to

Diagram Showing VPBoBB circuitcircuit

IN HOUSE DEVELOPMENT

Software Hardware

P wave detection

Parameters

Empirical relations

to predict size

Parameters

Warning dissemination

system

IN HOUSE DEVELOPMENT

Hardware

Siren

Sensor

Technical Aspect of Regional Warning System

P-Phase S-Phase & Surface waves

4 Stations (close to epicenter) pick the earthquake and send data to central server

Possible magnitude

is estimated

First 3 seconds of

record are analysed

Technical Aspect of Regional Warning System

Time (s)

earthquake and send data to central server

Possible magnitude

is estimated

If estimate is greater than

M6.0 warning is issued

OnSite and Regional EEW System

Threshold Limit for On

Shaking Starts Threshold Limit for On-Site Warning

Warning Issued to all Vulnerable

Start of Earthquake

at Epicenter

First

Second

Third

Fourth

Threshold Limit for On

Warning

Warning Issued to all Vulnerable

Locations by Regional EEW

Lead Time by

Regional EEW

and Regional EEW System

Threshold Limit for On-Site On-Site

Site Warning On-Site Warning Issued

S-Wave Threshold Limit for On-Site On-Site

Warning Issued

S-Wave

Arrival

Lead Time by On-Site EEW

Lead Time by

Regional EEW

*Maximum Lead time by On-Site

EEW for this case could be 25 secs

Algorithm

Cumulative Absolute Velocity

Bracketed Cumulative Absolute Velocity

Algorithm

Characteristic

Pd is maximum displacement

Predominant

Scaling relations developed for Garhwal Himalaya

Thresholds for different parameters for issuing warning (M

Scaling relations developed for Garhwal Himalaya

Thresholds for different parameters for issuing warning (M≥6)

Window(in sec)

τc (in sec)

τp(in sec)

Pd(in cm)

CAV(in cm/sec)

RSCV(in cm/sec)

1 1.02 0.95 0.13 3.0 0.3

2 1.17 1.00 0.27 8.0 1.0

3 1.20 1.06 0.51 10.0 1.7

4 1.42 1.10 0.95 23.0 5.2

5 1.55 1.14 1.38 41.0 10.0

Flowchart for decision making

Wave Ring

Monitor Wave Ring

for Event

Event

No

Data Archived for Future Reference

Pick Ring

Monitor Pick Ring4 or more

Picks found

No

Event

Detected

Yes

Flowchart for decision making

Data Archived for Future Reference

(SQL Server)

Estimate MPdYes

MPd > 6No Yes

EEW Warning System (Architecture)

Real Time

Sensor Data

Processing

Server

Sensor

Sensor BSNL VPN

& SWANServer

Siren

Sensor

Sensor

EEW Warning System (Architecture)

Warning

Server

Internet

Siren

Internet or

Radio Waves

Radio Siren Siren

Siren

Mock drills

� Training and mock drill are important part of any disaster

management system.

� We are preparing videos, posters and pamphlets for mock drills.� We are preparing videos, posters and pamphlets for mock drills.

� Mock drills and training will be provided to public by the

government agencies.

and mock drill are important part of any disaster

We are preparing videos, posters and pamphlets for mock drills.We are preparing videos, posters and pamphlets for mock drills.

Mock drills and training will be provided to public by the

Future work

� Development of more parameters and algorithm for avoiding false alarms

� Target dependent warnings – Earthquake magnitude and then Intensity at a specific point

� GIS based systems

� Hybrid networks containing on site and regional warning systems to work together

� Implementing it at least in Seismic Zone IV and V on National level

� Mock exercises with the help of sirens in EEW system.

� Development of mobile apps with some disaster management tips

� More use of RF and internet for awareness through these applications.

� Development of low cost indigenous sensors

Thanks & Acknowledgements….

Thanks to Ministry of Earth Sciences for funding this Pilot project.

Thanks to Uttarakhand Government for continuing this effort as disaster mitigation strategy.

� Development of low cost indigenous sensors

� Use of IOT on large scale implimentation

Development of more parameters and algorithm for avoiding false alarms

Earthquake magnitude and then Intensity at a specific point

Hybrid networks containing on site and regional warning systems to work together

Implementing it at least in Seismic Zone IV and V on National level

Mock exercises with the help of sirens in EEW system.

Development of mobile apps with some disaster management tips

More use of RF and internet for awareness through these applications.

Acknowledgements….

Thanks to Ministry of Earth Sciences for funding this Pilot project.

Government for continuing this effort as disaster mitigation strategy.

Thank you…..Thank you…..Thank you…..Thank you…..

EEW SensorsFirst Version of Sensor

� Specifications

� Tri axial MEMS Digital Accelerometer

� Dynamic range: 96 dB

� Recording range: +/- 2g

� Capability to calculate Tau C and Pd

� Capability to issue onsite warning

� Capability to stream acceleration time history to two different servers through LAN port

� Capability to receive warning from Central Station for issue of warning

EEW SensorsSecond Version of Sensor

Go Back

� Specifications

� Tri axial MEMS Digital Accelerometer

� Dynamic range: >100 dB

� Recording range: +/- 2g

� Capability to calculate Tau C and Pd

� Capability to issue onsite warning

� Capability to stream acceleration time history to two different servers through LAN port

� Capability to receive warning from Central Station for issue of warning

� Memory for offline data storage

P-onset

Thresholds for different parameters for issuing warning (M

Window(in sec)

τc (in sec)

τp(in sec)

Pd(in cm)

CAV(in cm/sec)

RSSCV(in cm/sec)(in sec)

c (in sec)

p(in sec)

d(in cm) (in cm/sec) (in cm/sec)

1 1.02 0.95 0.13 3.0

2 1.17 1.00 0.27 8.0

3 1.20 1.06 0.51 10.0

4 1.42 1.10 0.95 23.0

5 1.55 1.14 1.38 41.0

Go Back

Data streaming

from sensor

Data streaming

from sensor-01

STA

With HPF on CF

LTA

With LPF on CF

P wave detection

STA/LTA > TR(STA/LTA > TR(thr)

YES

NO

IF > 4 Sensors

YES

Data streaming

from sensor-04

Data streaming

from sensor-03

Data streaming

from sensor-02

LTA

With LPF on CF

P wave detection

Go Back

STA/LTA > TR(thr) STA/LTA > TR(thr)

STA/LTA > TR(thr)

YES

YES

P wave detection

Data streaming

from sensor-01

STA

With HPF on CF With LPF on CF

Data streaming

from sensor

STA/LTA > TR(thr)

YES

NO

IF > 4 Sensors

LTA

With LPF on CF

Data streaming

from sensor-02Data streaming

from sensor-03 Data streaming from sensor-04

STA/LTA > TR(thr) STA/LTA >

TR(thr) STA/LTA >

TR(thr)

Go Back

EEW Warning Server (Architecture)

Earthworm

Server

On a Big Earthquake confirmation warning

is disseminated to all sirens.

Warning messages are kept small to

reduce load on network.

Warning messages are encrypted.Server

Warning messages are encrypted.

EEW Warning Server also keep track of

sirens and in case of failure it should create

record and send notifications.

Database created for maintaining

information about Location, Current IP,

Current Status, Last Ping, Test Records of

sirens.

EEW Warning Server (Architecture)

Application

ServerWeb APIs

Earthworm

Server

Web Server

Internet

Go Back

EEW Warning Server

ServerWeb APIsServerInternet

Database

EEW Sirens

Encryption

Security Algorithm (Random Digit

Generator) 5 Digits

Random Number

Generator

Multiplied with pre

know number

ResultRandom Number

3 Digits Random Number

Middle 3 Digits

First Five Digits

Random Number

3 Digits Random Number Generator

Random Number

Designed an algorithm for keeping our communication

Security Algorithm (Random Digit

Truncated to 7 digits

3 Digits Random Number

Added with pre known number

Result

Middle 3 Last 7 Digits

Result

3 Digits Random Number Generator

Random Number

Designed an algorithm for keeping our communication secure Go Back

EEW Siren

Improved Version of Siren

Control Box Hooter

Three LEDs has been added for giving information about Power, System & Internet

Push buttons has been added for validating connections at the time installation,

testing of siren and creating easy test records

Different types of panel power connector has been added for making installation

easy and avoiding false connections

LAN PortPower SocketsPush Buttons

Audio Output

LED Lights

Siren

Hooter

First Version of SirenHooter

Go Back

HooterControl BoxRelay Circuit

Three LEDs has been added for giving information about Power, System & Internet

Push buttons has been added for validating connections at the time installation,

making installation

Improved Version of EEW Siren

Power Sockets

Internal View of Control Box

12-0-12V 2 Amp Transformer

Push Buttons

Amplifier Circuit

Improved Version of EEW Siren

Solid State Relay

Raspberry Pi

View of Control Box

LED Lights

12-0-12V 5Amp Transformer

5V Regulator Circuit

Sirens working on Radio Waves

Earthworm

Server

Warning

Server

Micro

Controller

RF

Transmitter

Relay

CircuitHooter

Micro

Controller Receiver

Sirens working on Radio Waves

Radio

RF

Transmitter

Radio

Waves

RF

Receiver

Transmitter

Receiver

Go Back

Sirens working on Internet

Raspberry

Pi

Relay

Circuit

Internet

SMPS

Sirens working on Internet

Hooter

• Internet connection using

• Lan Wire

• Wi-fi

• Internet Dongle

Go Back

EEW Siren Program Flow chart

Start

Turn on System LED

Co

nn

ec

ted

Co

nn

ec

ted

Turn on Internet LED

Send Info. to EEW Server

Internet

No

No

De

lay

30

Se

co

nd

s

Decrypt Code

AuthenticNoNo

Get Code

Turn off Internet LED

No

t R

ec

eiv

ed

No

t R

ec

eiv

ed

OKOK

EEW Siren Program Flow chart

Delay 60 Seconds

Relay Off

Decrypt Code

Extract Message

Relay On

Authentic

YesYes

Warning YesYes

De

lay

1

Se

co

nd

NoNo

Go Back

EEW Siren Interrupt Flow chart

Start on Test Button Press

Set PASS=True

Set TEST =TRUE

Start on PASS Button Press

SetTTIME=Curr. Time

END END

TEST==True

Set TEST =False

Blow Siren

FalseFalse

EEW Siren Interrupt Flow chart

Set PASS=True

Start on PASS Button Press

Set PASS=False

Start on FAIL Button Press

TTIME+5 mins > Curr.

Time

FalseFalse

Send Report to EEW Server

END

TEST==True

Set TEST =False

TrueTrue

TrueTrue

Go Back