Quantifying uncertainty in climate change impacts: A Bayesian approach for the Cagayan river basin

This study develops a probabilistic flood-impact assessment for the Cagayan river basin in northern Luzon, Philippines. Starting from the observation that climate change can intensify multi-day extreme rainfall and increase damaging riverine floods, the authors build a methodological chain linking Bayesian extreme-value inference with uncertainty propagation along the rainfall–flood–damage sequence. Precipitation data are modeled with Bayesian GEV distributions, 100-year return level design rainfall is estimated by combining three CMIP6 climate models through Bayesian model averaging, and hydrological response is simulated with the RRI model calibrated on a historical reference storm event.

The findings revealed that the SSP3–7.0 scenario significantly shifts extreme rainfall levels upward compared to the more optimistic SSP1–2.6 scenario, while simultaneously widening uncertainty in the estimates. This uncertainty propagates nonlinearly along the chain through to damages: the 95% credible range width for building damages nearly doubles from 1,944 to 3,817 million PHP, with similar increases for rice and corn losses. This demonstrates that deterministic estimates based on a single design level tend to understate tail risk under stronger climate forcing, making a probabilistic approach essential for robust adaptation planning and highlighting that reporting only central estimates can give a misleading picture of true exposure.