Interpreting the Causes of Summer River Disasters in Korea: An Analysis of Accident Investigation Reports

Recent advances in socio-technical systems have increased interdependencies among system components and created conditions in which the impacts of disasters can readily propagate across space and time. Consequently, contemporary disasters tend to be complex and large-scale, driven by multiple interacting causes rather than being explained by single triggering factors such as floods or landslides.

These characteristics emphasise the importance of identifying disaster causes in the pre-event phase to support prevention and preparedness strategies, beyond a sole focus on post-event response. In particular, the emergence of new risks associated with recent extreme weather events and climate change constrains the feasibility of reliable prediction, highlighting the need for systematic analysis of recurring causal factors and their interaction patterns observed in past disasters.

In response to these needs, causal analysis methods have evolved from first- to third-generation approaches. While first- and second-generation methods are effective in identifying single causes or failures of protective measures, they are limited in explaining modern disasters characterised by complex interactions among multiple actors and structural, environmental, institutional, and technological factors. By contrast, third-generation approaches consider system-wide control structures and non-linear causal relationships, enabling the analysis of preventive interventions even under conditions of uncertainty.

This study analyses 206 official accident investigation reports to examine recurring causal factors and damage amplification mechanisms. Representative causal analysis methods were selected for each generation, and the reports were analysed based on the key components required by each method. The results indicate that the extent to which essential components are addressed varies across causal analysis methods. In addition, structural and systemic causes account for a higher proportion than causes attributable to human error, and patterns of interaction among causal factors are found to be highly complex. Based on these findings, this study proposes an interpretive causal analysis framework for analysing the causes of summer river-related disasters in Korea, taking into account the limitations of single-cause–oriented analytical approaches.