Uncertainty quantification on risk inferences for compound hydroclimatic extremes

Multivariate risk analyses for compound hydroclimatic extremes (i.e., the simultaneous occurrence of multiple extremes) are challenged by extensive uncertainties arising from data availability, model selection, and parameterization. Neglecting these uncertainties can lead to unreliable risk inferences and ineffective resilience strategies. This project aims to develop innovative methodologies to account for and quantify these uncertainties in modeling compound hydroclimatic extremes. The research will also identify the dominant contributors to imprecise risk predictions, enhancing the reliability of risk assessments.

Key tasks include:

1) Development of copula-based Bayesian inference (CBI) approaches to robustly quantify model parameters and associated uncertainties.
2) Creation of a factorial-based uncertainty tracking system to identify the dominant contributors to uncertainties in risk inferences under varying climate scenarios.
3) Exploration of regional and combinatory variability, assessing how key uncertainty drivers differ across various compound extremes and geographical contexts.

The proposed methodologies will be applied to diverse combinations of extremes across multiple regions, enabling more accurate and actionable risk predictions. The findings will provide critical insights to support targeted, efficient mitigation policies, contributing to global resilience against climate-related compound risks.