Measuring Seismic Risk in Turkey - Latest Modeling Techniques

1
Measuring Seismic Risk in Turkey - Latest
Modeling Techniques

• Introduction and moderation
• Anselm Smolka, Munich Re
• Panelists
• Dennis Kuzak, EQECAT
• Oliver Peterken, Willis
• Fouad Bendimerad, RMS

2
The Istanbul earthquake of 1509
3
Three models and three answers why the
differences?

• Turkey-specific factors
• General modelling related issues

4
Risk models general approach
5
Turkey-specific factors - Hazard

• Seismicity disagreement about Marmara region
• Source zones/ fault geometry
• Time dependence
• Collateral hazards
• which ones considered, and how ?
• Subsoil conditions
• data base
• assigning average class

6
General modelling issues

• Ground motion parameter
• Macroseismic MMI/EMS
• Spectral acceleration
• Displacement response spectrum
• Attenuation does not seem an issue for Turkey,
• but magnitude threshold for fault length

7
General modelling issues

• Vulnerability
• Difficult comparison for different g/m parameters
• threshold value for beginning damagegt great
  influence on AAL
• need for testing against actual loss
  experience(I dont believe in statistics but the
  ones falsified by myself...)

8
Conclusions

• Putting together best components does not yet
  mean a good model
• Final results must be tested against actual
  recent and significant historic events
• In terms of statistical stability the number of
  simulations is important
• Wi(l)dely disagreeing results between models are
  not a good service to the market