Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Natural Disaster In Canada (2024)

H. Hao

TL;DR

This paper analyzes the Canadian Disaster Database to quantify how often meteorological and hydrological natural disasters occur in Canada, how severe the losses are, and how national annual losses may be shifting. It uses a two-stage Poisson–GPD framework to model per-type annual losses and employs Fast Fourier Transform (FFT) to compute univariate loss distributions, with a Normal copula to aggregate across disaster types. The results indicate no strong nationwide increase in disaster frequency over the last seven years, but significant growth in tail losses for Thunderstorm and Wildfire, consistent with warmer, wetter climate conditions and implying a higher likelihood of extreme national losses. The approach delivers efficient, scalable loss-distribution estimates for multi-type hazard portfolios, informing risk management and policy planning in Canada.

Abstract

This paper is a follow-up to our earlier study, Natural Disasters in Canada (2017). We analyze the Canadian Disaster Database (CDD) to examine the frequency and severity of various natural disasters over the past 120 years and to identify emerging trends. We generate annual loss distributions for individual disaster types, as well as an aggregate annual loss distribution across all event types. Our analysis provides evidence that Canada is experiencing warmer and wetter conditions and indicates a substantial likelihood of extreme national-level losses.

Natural Disaster In Canada (2024)

TL;DR

This paper analyzes the Canadian Disaster Database to quantify how often meteorological and hydrological natural disasters occur in Canada, how severe the losses are, and how national annual losses may be shifting. It uses a two-stage Poisson–GPD framework to model per-type annual losses and employs Fast Fourier Transform (FFT) to compute univariate loss distributions, with a Normal copula to aggregate across disaster types. The results indicate no strong nationwide increase in disaster frequency over the last seven years, but significant growth in tail losses for Thunderstorm and Wildfire, consistent with warmer, wetter climate conditions and implying a higher likelihood of extreme national losses. The approach delivers efficient, scalable loss-distribution estimates for multi-type hazard portfolios, informing risk management and policy planning in Canada.

Abstract

This paper is a follow-up to our earlier study, Natural Disasters in Canada (2017). We analyze the Canadian Disaster Database (CDD) to examine the frequency and severity of various natural disasters over the past 120 years and to identify emerging trends. We generate annual loss distributions for individual disaster types, as well as an aggregate annual loss distribution across all event types. Our analysis provides evidence that Canada is experiencing warmer and wetter conditions and indicates a substantial likelihood of extreme national-level losses.

Paper Structure

This paper contains 16 sections, 7 equations, 4 figures, 6 tables.

Table of Contents

  1. Summary
  2. Trend
  3. Frequency and Severity
  4. Modelling of Frequency
  5. Modeling of Severity
  6. Simulation of Annual Losses for Each Disaster Type and Their Aggregate
  7. Simulation of Annual Losses for Each Disaster Type
  8. Aggregation of Annual Losses for All Disaster Types
  9. Conclusion
  10. Natural Disasters Database
  11. Mathematical Background
  12. Overview
  13. Model Setup
  14. Characteristic Function of Total Loss $S$
  15. FFT Procedure
  16. ...and 1 more sections

Figures (4)

  • Figure 1: Number of Disasters since 1900
  • Figure 2: Losses from Disasters since 1900
  • Figure 5: Annual Loss Simulation by Event Type: 2024 Collection
  • Figure 6: Simulated and Realized Annual Total Losses Across All Event Types