PhD Thesis Presentation - A Systematic Approach to Identify Unknown Unknowns for Early Warning in the Era of Climate Change: Implications for Proactive Climate Governance

Climate change is not only increasing the frequency and intensity of well-known hazards, such as floods and heatwaves, but also the emergence of complex impacts that fall outside the scope of conventional risk management systems. These emerging risks are often subtle at first, followed by the recurrence of similar tragedies across regions. For instance, after four people drowned in flooded underground car parks in Macau in 2017, similar tragedies killed seven in South Korea (2022) and over 200 in Spain in (2024). This recurrence reveals a common pattern: conventional risk management assumes climate risks are stationary, while climate change creates a nonstationary environment with changing risk patterns. Consequently, conventional risk management is inadequate for emerging threats, forcing significant policy reform only after a local catastrophe. Previous research underscored the significance of ‘unknown unknowns’—rare, variable-severity emerging risks—and highlighted a clear need for new identification methods. But it lacked practical detection methods. Addressing this gap, this study demonstrates a systematic approach that leverages cross-regional knowledge of analogous events to identify ‘unknown unknowns’ for regions without prior experience, transforming them into foreseeable risks. We employ Natural Language Processing to analyze 7.7 million news articles, identifying 639 impacts. After being refined through expert intervention, the findings are classified into two types of emerging threats. The first category includes previously unexpected risks, whose connection to climate change is surprising and not obvious. The second category involves lesser-known risks with an understandable connection to climate change, but these risks are not yet covered by the existing risk management systems. Ultimately, this research delivers a systematic methodology for translating these findings into actionable insights for decision-makers. It enables a shift in climate governance from post-tragedy analysis to proactive prevention, supporting adaptive climate policies, aligning with the UN’s ‘Early Warnings for All’ initiative.