A Multisource Information System in support of Early Warning Systems in Brazil: The REINDESC database

Systematic documentation of disaster events is fundamental for supporting early warning systems (EWS), particularly in establishing critical thresholds and analyzing spatial patterns. Nevertheless, low-severity floods and landslides remain largely underreported because their impacts are limited and often overlooked. This study presents a comprehensive, multiscale, multisource information platform developed to strengthen early warning capacities in Brazil. The database’s central purpose is to address the nationwide absence of systematic records for small-scale hydrogeological hazards. We present the REINDESC database, describe its methodological framework, and highlight its relevance to the Brazilian EWS. Data from multiple sources were compiled comprehensively, and events were systematically classified by magnitude, temporal accuracy, and spatial precision. Between 2016 and 2023, more than 11,000 occurrences were documented, capturing the spatial and temporal patterns of low-magnitude, high-frequency events. During this period, nearly 7 million individuals (6,961,199) were affected by intensive events associated with intermediate- and high-magnitude occurrences. In contrast, almost 2 million people (1,961,322), representing 22% of the total affected population, were affected by extensive events associated with low-magnitude occurrences. Southeastern Brazil was the most affected region, with nearly 3 million residents impacted, of whom 41% experienced extensive-risk events. The findings underscore the cumulative significance of low-magnitude hydrogeological events in Brazil and highlight the need to integrate such occurrences into national risk reduction strategies. Based on this experience in developing collection procedures, we identified several advantages and challenges, limitations, and gaps in consolidating a disaster database. REINDESC demonstrates substantial potential for strengthening EWS by providing a robust, standardized, and spatially explicit record of hydrogeological hazards across the country. Identification and classification of high frequency/low risk events and low frequency/high-risk disasters is a valuable resource for the Brazilian Early Warning System. Highlighting the impacts of both small and large disaster events underscores their importance in regions with a higher frequency of hydrological and geological events. • The REINDESC is an innovative database that catalogues hydrogeological events in Brazil, encompassing not only major disasters but also small-scale floods and landslides. • Given that small-scale events occur frequently, REINDESC emphasizes their significance due to their cumulative impacts. • The database utilizes a wide array of sources, ensuring comprehensive information with regard to accuracy, depth, reliability, and trustworthiness. • REINDESC is designed for rapid data acquisition, providing information on landslides and floods within hours of their occurrence and offering timely updates. • The applications of REINDESC have significantly enhanced Brazil's early warning system, making it essential for assessing and verifying flood and landslide warnings.