Quantifying Physical Climate Risks for Key Industrial Sectors in Taiwan: A Financial Impact Assessment of Flood Hazards under Multiple Climate Scenarios

Climate change is expected to intensify extreme precipitation, increasing future flood-related losses. Yet, prioritizing adaptation remains challenging without credible estimates of the financial impacts of physical climate risk. This study develops an integrated analytical framework to quantify flood-induced financial losses in Taiwan, specifically focusing on the semiconductor, cement, petrochemical, and steel industries. The framework translates climate-driven hazard changes into asset-level value impacts for these critical industrial facilities.

The methodology integrates historical station observations with statistically downscaled precipitation projections from AR6 GCMs. Future daily rainfall is simulated using a multi-site stochastic weather generator (MultiWG). These series are then disaggregated to hourly rainfall using a feature-vector-based k-nearest neighbors (KNN) resampling approach. While general scenarios rely on GCM simulations, this study augments the stress testing framework with bias-corrected AR5 typhoon dynamic downscaling data to better capture extreme event dynamics at higher spatial resolutions. To bridge the gap between rainfall and flood impacts, ten temporal patterns from Taiwan’s Water Resources Agency (WRA) are utilized to estimate scenario-specific frequencies of extreme rainfall. Inverse distance weighting (IDW) is subsequently applied to interpolate location-specific extreme-rainfall frequencies to estimate localized inundation depths based on WRA flood potential maps. WRA depth–damage curves are then overlaid to estimate expected asset losses over 20-year horizons for a historical baseline (1995–2015) and three future periods (2021–2040, 2041–2060, and 2061–2080) under multiple climate scenarios.

Rather than focusing on absolute financial loss figures, this study emphasizes a comparative analysis of average annual losses and tail-risk impacts, quantified through Value-at-Risk (VaR), across the selected industrial sectors. By mapping these quantified risks onto financial statement line items, the framework supports decision-useful reporting and evaluates system stability under extreme events through climate stress testing. Ultimately, this framework facilitates sensitivity analysis to identify priority adaptation targets and optimize investment portfolios. These outputs strengthen TCFD-aligned disclosure by offering a transparent and defensible basis for communicating physical risk and adaptation actions in the industrial sector.