Workshop on the Interpretation of Earthquake

Risk Assessment Data

Aqaba, 12-14 July, 2010

Dr. Hanan Al-Nimry

Jordan University of Science and Technology
According to Dowrick (1987) Seismic Risk can be
defined as:

Seismic Risk = (SeismicHazard) x

(Vulnerability) x (Value)

where:
Vulnerability is the amount of damage induced by a
given degree of hazard, and expressed as a ratio of
the Value of the damage to the total cost of the item
under consideration.
Earthquake Risk Assessment
Hays (1994), suggested that the determination of
the seismic risk requires four elements. These
are :

•The temporal and spatial characteristics of the

hazards,
•The exposure (i.e. inventory of buildings and
lifeline systems),
•The location of buildings and lifelines relative to
the disaster agents, and
•The vulnerability (i.e. expected loss of value or
function)
the Seismic Risk has also been defined as the
convolution of three factors:

•Seismic Hazard which includes all direct

physical effects produced by an earthquake;
•Induced Physical Hazard, refers to those events,
like landslides, that might be triggered by ground
motion; and
•Systemic Vulnerability, refers to the
vulnerability of urban and regional systems
which cannot be considered only as the sum of
the vulnerability of exposed structures
SEISMIC HAZARD

The probability of seismic shaking reaching or

exceeding a certain magnitude (or macroseismic
intensity) during a reference period determines
the Seismic Hazard for a
particular area.

Other parameters, which can also determine the

Seismic Hazard, are maximum acceleration,
velocity or displacement
EVALUATION OF SEISMIC HAZARD

Seismic Hazard Assessment (i.e. identification

of the spatial and temporal distribution of the
disaster agent) can be made on national,
regional, urban and site specific scales.
Scale of investigation (after Hays, 1994)
The potential phenomena generated by an
earthquake are ground shaking,
liquefaction, landslides, tectonic
deformation, surface fault rapture,
tsunamis, release of toxic materials, fire
and aftershocks.
There are various methods that have been
used to assess the Seismic Hazard for a
given region:

1. Observational
2. Deterministic
3. The Statistical
4. Probabilistic Seismic Hazard Assessment
(PSHA)
5. Time dependent
insurance companies have used the PSHA method,
in an attempt to calculate appropriate premium
rates for earthquake insurance or to determine
the range of probable maximum loss that would be
incurred in the event of a catastrophic
earthquake.
PROBABILISTIC SEISMIC HAZARD
ASSESSMENT (PSHA)

PSHA methods involve probabilistic modelling and

statistical analysis. These methods were first
developed in the late 1960s (Cornell, 1968, in
Grunthal, et al., 1995; Cornell, 1971) and are now
used regularly.
The probabilistic approach proposed by Cornell
(1968) includes the following steps:

1. The geometrical identification of seismic sources

(i.e. define seismogenic units either area source
zones or individual faults);
2. The development of models describing the
seismicity characteristics of the sources (i.e. quantify
their magnitude recurrence (M-N) relation,
earthquake activity);
3. The determination of attenuation laws for the
relevant parameters between the sources and the
site (i.e. quantify their Intensity (I) -Distance
relation, or peak ground acceleration (PGA) (α) -
Distance relation);
4. The addition of the contribution of all the
The main aim of this method is to estimate the
probability that a particular relevant parameter
(i.e. acceleration or intensity) will be exceeded
during a specified period of time at the site under
consideration.

By estimating the probabilities of the other

intensity values or PGA, for various time periods,
a complete account of the hazard at the specific
site can be obtained.

One of the main difficulties of the PSHA method is

the definition of the initial seismic source zone
model. This method is suitable for areas with
well-documented seismic history and sufficiently
recorded seismic activity.
The accuracy of the predicted hazard at the
site will depend upon the quality of the input
information (i.e. the configuration of sources and
their activity rates, the form and parameters of
the attenuation relationship and the treatment of
local effects).

Small differences in the model can have quite

large effects on the hazard. It can be concluded,
therefore, that the features of the input
parameters are more critical than the
methodology of the calculation of the
seismic hazard.
The PSHA method was initially developed
particularly for individual sites, but it can be
uniformly applied for a grid of points to obtain a
regional seismic map with a certain probability
level. In that case the values are contoured to
illustrate the spatial variation of the hazard.
All the parameters involved in the estimation of
seismic risk can be given geographical co-
ordinates and displayed and analysed in the form
of maps.

A Geographical Information System (GIS) is a

computer system able to efficiently assemble,
store, manipulate, and display geographically
referenced information (data identified according
to their location).
For the assessment of the earthquake risk to be complete it
must include an assessment of the probable levels of human
casualties, both deaths and injuries resulting from an
earthquake.

The loss of life during an earthquake is caused in many

different ways such as:
• Building collapse or collapse of building components or
contents
• Machinery accidents occurring during the disturbance
• Medical conditions induced by the shock of experiencing
ground motion (i.e. heart attacks)

In addition to these, some earthquakes trigger follow-on

hazards, such as rockfalls, landslides, mudflows, tsunamis
and fires, which create further loss of life. Furthermore,
epidemics among the homeless or even shootings during
martial law are also included in the list of life-loss hazards.
However, the main cause of fatalities due to earthquakes is
structural failure or partial collapse.
On the other hand, in cases where higher
quality of building stock is involved (low
vulnerability), such as in the USA or
Japan, the biggest numbers of fatalities
are caused by the failure of non-structural
elements (i.e. dislodged pieces from the
exterior of buildings, collapse of free
standing walls).
Coburn and Spence (1992) suggest that the total number
of deaths caused by an earthquake can be expressed as:

where:
KS is the number of fatalities due to structural damage
K′ is the number of fatalities due to non-structural
damage
K2 is the number of fatalities arising to follow –on
hazards (rockfalls, landslides, mudflows, tsunamis,
fires)
LETHALITY RATIO

According to Coburn and Spence (1992) the

lethality ratio is the required parameter and is
basically the relationship between the number of
people killed and the number of collapsed
buildings. If this ratio is known, then from the
estimates of the collapsed buildings it is possible
to estimate the human casualties.
The equation below expresses the number of fatalities,
resulting from the damage caused to the specific class
“b” buildings.

where:

D5b is the total number of collapsed structures (damage

level 5 i.e. for load bearing masonry - more than one wall
collapsed or more than half of roof, for RC framed
buildings – failure of structural members to allow fall of
roof or slab) of buildings of type b

M1 to M5 are a range of modifiers to a potential

mortality figure
M1 – Population per Building (after Coburn and Spence, 1992)
M2 – Occupancy at time of earthquake (after Coburn and
Spence, 1992)
M3 – Occupants trapped by collapse (after Coburn and
Spence, 1992)
M4 – Injury distribution at collapse (after Coburn and
Spence, 1992)
Example:
Earthquake Risk Assessment (ERA) model
(EQ-RACY: EarthQuake Risk
Assessment for Cyprus),

EARTHQUAKE RISK ASSESSMENT

& MANAGEMENT CASE STUDY: CYPRUS
A thesis submitted for the degree of Doctor
of Philosophy in the Faculty of Engineering
of the University of Sheffield
By STELLA KYTHREOTI
PART I – SEISMIC HAZARD ASSESSMENT

Part I of the model deals with the estimation of

the seismic hazard. The three input databases of
Part I are listed in Table 4.1. The routine starts
by calculating the epicentral distance (Rij) for a
particular location (Vi) from a particular
earthquake (Qj) (using the Pythagorean theorem)
and then it estimates the peak ground
acceleration (αij) and finally the intensity (Iij)
through a set of attenuation laws.
PART II – EARTHQUAKE RISK ASSESSMENT

Part II of the model involves the ERA. By checking

all individual intensities (for each earthquake)
greater or equal to V the model uses the
vulnerability models to calculate damage. If the
intensity for a particular location and event is less
than V, the model terminates the process, since
low intensities do not cause any significant
damages.
PART III – FATALITIES AND INJURIES

This part deals with the estimation of the fatalities

and injuries (typical maps shown in Figure 4.10).
This is achieved by the use of the Mean Fatalities
Ratios (MFR) and the Mean Injuries Ratios (MIR),
the population presence that is dependent on the
time of occurrence (i.e. month of year and hour of
day) of the events as well as the population
distribution.
RISK MANAGEMENT

the reduction of community earthquake

vulnerability is a long-term process, which
involves the creation of a basis for changing the
way the community assesses and evaluates its
hazard, built, and policy environments.

Every community can reduce its vulnerability by

taking the appropriate actions that can either
prevent or minimise the physical and social
demands of earthquakes and increase the
community’s ability to cope with them. A number
of actions that can be considered part of a
community’s options for risk management,
suggested by Hays (1994), are presented in Table