“An Early Earthquake Loss Assessment Framework” By: Shaukat …solutions.ait.ac.th/wp-content/uploads/2017/05/SK-An... · 2017. 5. 4. · Seismic Hazard Assessment Industrial

04/05/2017 © The University of Sheffield

“An Early Earthquake Loss Assessment

Framework”

By: Shaukat Khan


An Earthquake Risk Assessment (ERA) framework based

on GIS is under development in an ongoing research at the

University of Sheffield, UK.

The ERA framework uses a new Probabilistic Seismic

Hazard Assessment method.

The aim of this research is to enhance the ERA framework

with an early earthquake loss assessment component.

Seismic Hazard

Assessment

Industrial

Vulnerability

Risk

Assessment

Casualty

Assessment

Instrumental

Seismicity

Building

Vulnerability

Historical

Seismicity

Faults and

Gaps

Building

Characteristics

Industrial

Components

Industrial

Inventory

Building

Vulnerability

Relationships

Vulnerability

Relationships

for Industrial

Components

Population

Properties

Casualty

ParametersFatality Risk

Injury Risk

Monetary

Risk

PSHA

Tsunami Hazard

Assessment PTHA

Building

Inventory

Ha

za

rd M

od

ule

Vu

lne

rab

ility

Mo

du

leR

isk M

od

ule


ERA

Framework

RISK/LOSS = HAZARD X VULNERABILITY X VALUE


A new Probabilistic Seismic Hazard

Assessment (PSHA) methodology

was developed for the Hazard

component of the ERA framework

To avoid the problems in existing PSHA

methodology


Existing PSHA methodology

Minimum magnitude selection (hazard underestimation).

Maximum magnitude selection (hazard overestimation).

Use of recurrence relationships (inaccuracy).

Smearing of smoothing of seismicity over source zones.

Defining of seismic source zones is subjective.

Requirement of good knowledge of historical seismicity

and tectonic setting which are not readily available in

developing countries.


New PSHA methodology

Site PGA cut off is used instead of minimum magnitude

Maximum magnitude is not used (for ERA studies)

Actual data is used instead of determining recurrence

relationships

Actual seismicity is used (smoothing not required)

Good results with crude seismic source zones

Minimal knowledge of historical seismicity and tectonic

setting can be used


Future seismicityHistorical seismicity

Instrumental seismicity

New PSHA process


Synthetic catalogues using

Monte-Carlo simulations

Synthetic

catalogues

PGA

calculation




PSHA

Probability

calculation


Dealing with Large Meg. Events

Focal point of

new event

Isoseismals

New event location is

randomised within

this area

Site of

interest

Original

earthquake event

New EFL

Line at 30 km from EFL

Known fault

Focal point of new event

Isoseismals

(a) (b)


PGA calculated

from nearest

point on EFL

Original event

Known fault

line

Area within which

new event can take

place

EFL

New random

event

EFL of

new event

12.5 km cut

off on PGA

New random event for large magnitude earthquake shown with PGA distribution

Hazard Parameters

Magnitude

Location

Depth

Geology /soil parameters

Original event

Known fault

line

Area within which

new event can take

place

EFL

New random

event

New random event for large magnitude earthquake shown with PGA distribution

Large magnitude in Chile (2010) earthquake shown with PGA distribution


PSHA Map for Pakistan, 50 year return period with 10% POE

Current PSHA Methodology BCP (2007) using EZ-FRISKTM PMD and NORSAR (2007) using CRISISTM


ERA for study region in Pakistan


Vulnerability Component


SPATIAL DISTRIBUTION OF EXPOSED BUILDINGS (BUILDING INVENTORY)


Remote

sensingMinimal data collection from

remote sensing and field

surveys

Area of interest


Annual ERA

Probability

calculation

Loss

calculation


Synthetic catalogues using

Monte-Carlo simulations

Synthetic

catalogues

PGA

calculation



PSHA


Number of Deaths with time after an EventCoburn and Spence (2004)

No. o

f D

eath

s

Time in Hours

24 hrs.


Kashmir (2005) earthquake

The death toll was around 3,000 in the first few hours of the event

The death toll increased to more than 25,000 in the first few days

The final number of deaths was more than 85,000 a few weeks after the event


Early Earthquake Loss Assessmentfor Large Catastrophic Events

An Early Loss Assessment component is to be added to the ERA framework

To generate PGA map for an event, it uses Monte Carlo simulations to generate randomised events from the data available from networks like the USGS

Event parameters like magnitude, depth and location are randomised and tectonic, infrastructure and population information are made available beforehand for the process


Magnitude

Depth

Longitude

Latitude

Event

Parameters

-m to +m

-20 to +20 Km

-Lon to +Lon Km

-Lat to + Lat Km

Range of Parameteric

ErrorsGenerate Randomized

events

List of Randomized Events

Calculate fault length, its orientation and distances of each site from each

randomized event in List

Seismic Zonation

Table of distances for all sites

Calculate PGA for all sites due to all randomized

Events in the list and select PGA value against given

PONE

Site Soil condition

PGA Values for Each Site

Spatial Plot of PGA values PGA Map


Legend (g)

a b

c

Earthquake

Mw= 7.2 near

the

Baluchistan

Arc, on

January 18,

2011

Processing

Time = 3 hrs

(after

determining

event data)


Earthquake

Mw= 7.6 in

Kashmir, on

October 8,

2005

Processing

Time = 3

hrs

Legend(g)

c

a b


Early Loss Assessment Framework Under Development

Integrating the early PGA component to ERA framework will make it possible to generate Injury and Fatality distribution for an event within the first 8 hours after its occurrence.


Conclusions

An ERA framework based on GIS is developed

An early earthquake shaking effect assessment tool is developed that can be incorporated into the ERA framework for early loss and casualty assessments


Thank you

Documents

“An Early Earthquake Loss Assessment Framework” By: Shaukat …solutions.ait.ac.th/wp-content/uploads/2017/05/SK-An... · 2017. 5. 4. · Seismic Hazard Assessment Industrial