Avaliação da eficácia de alertas para deslizamentos em Jaboatão dos Guararapes (PE)

Introduction. Landslides occur when slope stability is compromised by natural factors, such as intense rainfall, and human factors, such as irregular occupation. Climate variations have made extreme precipitation events more frequent and severe, increasing the social, economic, and environmental impacts of disasters. Objective. This study evaluates the effectiveness of the Automatic Rainfall Half-Life Cluster System (ARHCS) for monitoring landslides in Jaboatão dos Guararapes comparing its alerts to those issued by National Center for Monitoring and Early Warning of Natural Disasters (Centro Nacional de Monitoramento e Alertas de Desastres Naturais, Cemaden). Methodology. ARHCS uses the concept of rainfall half-life to estimate soil saturation potential and incorporates cluster analysis and data inversion to define alert thresholds. The comparison with Cemaden’s alerts was conducted for the year 2021. Results. The results indicated high compatibility between the systems, with ARHCS demonstrating greater accuracy in issuing alerts for “high” and “very high” risk levels, in addition to a low incidence of alerts without occurrences. Conclusion. These findings suggest that ARHCS can complement the existing alert system, especially in municipalities with limited monitoring infrastructure.