Dominant barriers and social injustices in flood early warning systems governance: insights from topic modeling and content analysis

Introduction Effective flood early warning systems (FEWS) are crucial to mitigating flood impacts. Yet, their governance is often hindered by numerous systemic barriers. These systemic barriers reinforce social inequities, which in turn exacerbate flood risk management and hinder the achievement of disaster justice. Existing studies provide limited insights into the dominant barriers that most significantly hinder effective governance of FEWS. Moreover, research examining the underlying social injustices embedded in these barriers remains scarce, despite the intensifying impacts of floods on existing social inequities, with the most severe consequences affecting vulnerable populations. Methods In this study, we employed a Latent Dirichlet Analysis (LDA) topic modeling to identify the dominant barriers to FEWS governance, followed by a qualitative content analysis to examine the underlying social injustice dimensions within these barriers. Results and discussion The topic modeling identified five interpretable dominant barriers per FEWS component from the corpus. The analysis revealed that capability injustice is the most prevalent primary social injustice underpinning the identified key barriers, followed by distributional injustice. Capability injustice, the unequal opportunties to utilize or access opportunities to enhance capabilities to operate or benefit from the system, remains an overarching dimension of injustice that is intensified by other forms of injustice. The study emphasizes the need for social justice to be considered in FEWS policy and governance. Going forward, addressing FEWS governance barriers requires attention to the underlying social injustices that perpetuate them, in order to avoid the risk of reinforcing existing inequalities within the system.