Assessment of seismic loss dependence using copula

Area of Science:

• Civil Engineering
• Risk Management
• Statistical Modeling

Background:

• Seismic risk is amplified by spatiotemporal correlations in building and infrastructure losses.
• Accurate assessment of aggregate seismic losses is crucial for effective risk management.
• Understanding the upper tail behavior of loss distributions is particularly important.

Purpose of the Study:

• To investigate seismic loss dependence between closely located building portfolios.
• To develop a statistical model for seismic loss dependence incorporating spatiotemporal correlations.
• To evaluate the accuracy of the proposed model for risk assessment.

Main Methods:

• Simulated seismic loss samples from a risk model with correlated ground motions.
• A loss frequency model with a common dependent random component.
• A copula-based loss severity model with upper tail dependence.

Main Results:

• The dependence structure of aggregate seismic losses can be effectively modeled using right heavy tail or Gumbel copulas.
• The proposed method demonstrates satisfactory accuracy for practical probability levels (≤10% error).
• The model is applicable to wood-frame buildings in regions like southwestern British Columbia.

Conclusions:

• The developed statistical seismic loss model accurately captures dependence structures.
• The model can enhance seismic risk assessment and financial analysis.
• This approach offers a reliable method for evaluating aggregate seismic losses.