Critical threshold curves for landslides in Rio de Janeiro city via long term rainfall monitoring

The city of Rio de Janeiro, Brazil, regularly experiences landslides triggered by rainfall. To enhance our understanding of critical thresholds, this study analyses two datasets over a period of 24 years (1999 to 2022): rainfall events, from data obtained from 33 city-installed rain gauges and an inventory of landslide occurrence reports from the Geo-Rio Foundation. The dataset comprises 7,019 significant rainfall events, of which 6,481 resulted in landslides. A total of 6,671 landslides occurred during the analyzed period, a total of 14,402 technical inspections were conducted after slope instability assessments were requested by the population. In addition to evaluating rainfall values that lead to landslides, the study analyses the impact of Geo-Rio’s slope-stabilization efforts and population growth in landslide-prone areas. Computational techniques were used to manipulate and classify the data, revealing zone-dependent thresholds. A strong correlation is identified between landslide events, rainfall intensity and accumulated rainfall (96-hour period). The results underscore the necessity to improve Rio de Janeiro’s early-warning system, considering both rainfall intensity and 96 hours of accumulated rainfall data to predict landslides, with implications for future slope stabilization efforts and urban planning in landslide-prone areas.