Use of Paleoseismic Data in Seismic Hazard

Analysis: Examples from Europe

Use of Paleoseismic Data in Seismic Hazard

Analysis: Examples from Europe

K.Atakan and A.Ojeda

Institute of Solid Earth PhysicsUniversity of Bergen

Allégt.41, N-5007 Bergen, NorwayTel: +47-55-583413 Fax: +47-55-589669

E-post: atakan@ifjf.uib.no

ProbabilisticSeismic Hazard

Assessment

SeismicRisk

Class

Historicalseismicity

Activefault data

InstrumentalSeismicity

Building parameters

Site parameters

RegionalSeismic Hazard

Earthquake sources

Seismic Wave Attenuation

Earthquake RecurrenceModels

Geodetic/ geophysical

data

Local Site Effects

Site-specificSeismic Hazard Seismic

Risk Index

Ambraseys

Boore et al.

Campbell

Sadigh et al.

Model 3

Model 2

Model 1

Earthquake Engineering

Geotechnical Engineering

Earthquake Seismology

Geology & Geophysics

Civil Engineering

Vulnerability of the site

Site-specific Spectral Hazard

Seismic Hazard and Risk

(McCalpin and Nelson, 1996)

(McCalpin, 1996)

Probabilistic

Seismic Hazard Assessment

Deterministic

Site specificanalysis

ScenarioBased models

PoissonianModels

RenewalModels

Seismic Hazard Assessment

HybridModels

Seismic Hazard in the Catalan Coastal Ranges

Seismic Hazard in the Catalan Coastal Ranges

Probabilisitic seismic hazard maps for various return periods. Earthquake occurrence is based on a Poissonian model.

Seismic Hazard in the Catalan Coastal Ranges

0-1

ELAPSED TIME1 RI 2 RI

CO

ND

ITIO

NA

L P

RO

BA

BIL

ITY

Poissonian

Renewal

M2

M4

M3

M1

Seismic Hazard in the Catalan Coastal Ranges

0

0

Mod

els

Present

TIME

85%

10%

0%

RI

RI

RIM2

M3

M4

Seismic Hazard in the Catalan Coastal Ranges

Maps showing the difference between the renewal and the poissonian models at various return periods in terms of PGA (in cm/sec2). The influence of the paleoseismic data become more visible at larger return periods.

UncertaintiesUncertainties

In General:uncertainties may be divided in two categories:

•Uncertainties due to the lack of sufficient data these can be improved with additional data

•Uncertainties due to the lack of understanding of the phenomena

additional data may not necessarily improve the understanding

In Paleoseismology:uncertainties may be grouped into two:

•Analytical and/or numerical uncertainties•Uncertainties related to the interpretation of dataand/or observations

Preferred sequence of paleoseismic investigationsPreferred sequence of paleoseismic investigations

•Regional Scale (thousands of km2)•remote sensing, aerial photo’s, •geological mapping, geophysical investigations•and other background knowledge

•Local Scale (a few km2)•geomorphic mapping, Quaternary •stratigraphic framework

•Site Scale (1 hectare to a few m2)•geophysics (shallow depth/high resolution)•fault-zone trenching•other detailed observations and•data collection

•Stage 1: Regional scale investigations that are dependent on

the rate of deformation/tectonic setting and the background

knowledge•Stage 2: Local scale investigations and the site selection for

detailed analysis•Stage 3: Extrapolation of the observations that are made at a

site scale to the entire fault. •Stage 4: Identification of the paleoearthquake(s) based on

the diagnostic criteria•Stage 5: Dating techniques used for the age determination of

the paleoearthquake(s)•Stage 6: Paleoearthquake size estimate and the recurrence

interval

Paleoseismological interpretation processgoes through the following stages:

UNIPAS acts as a link between the paleoseismic data and

seismic hazard assessment

UNIPAS acts as a link between the paleoseismic data and

seismic hazard assessment

Paleoseismic data

UNIPAS

Seismic hazardassessment

Paleoseismic Quality FactorPaleoseismic Quality Factor

The logic-tree analysis used in the paleoseismic interpretation process may be interfaced with the logic-tree analysis in the seismic hazard assessmentsthrough Paleoseismic Quality Factor (PQF). PQF isexpressed by the following:

PQF = Pes x Cri

where, Pes is the probability of the preferred end-solutionin the logic-tree analysis for the paleoseismic investigationand Cri is a correction term for the relative level of importanceof the investigation in seismic hazard analysis.

Level of importance of the paleoseismic investigation in seismic hazard assessmentLevel of importance of the paleoseismic investigation in seismic hazard assessment

The relative levels of importance may be grouped into five categories:

Level 1: Site-specific seismic hazard analysis (SHA)Level 2: Regional probabilistic seismic hazard assessmentLevel 3: Input as seismotectonic sources in probabilistic SHALevel 4: Identifying the earthquake potential of the fault (zone)Level 5: Determining if the fault (zone) is active (i.e. observable co-seismic slip during the Holocene)

PQF is the connecting link between the logic-tree for paleoseismic data and the

logic-tree for seismic hazard assessment

PQF is the connecting link between the logic-tree for paleoseismic data and the

logic-tree for seismic hazard assessment

Logic-tree forpaleoseismic

data

PQF

Logic-tree forseismic hazard

assessment

(McCalpin and Nelson, 1996)

(Wells and Coppersmith, 1994)