Descripción:
Floods are among the most destructive natural disasters, affecting millions of people worldwide and causing severe humanitarian, economic, and social impacts. This study synthesises available research on the application of Machine Learning (ML) for flood forecasting, directly supporting Sustainable Development Goal SDG 11 (Sustainable Cities and Communities). Furthermore, it aligns with the Sendai Framework’s 2015–2030 mandate to leverage advanced technologies for proactive disaster risk reduction. To do that, we conducted a systematic literature review (tertiary review) of 35 secondary studies published between 2015 and 2025. Our analysis identifies and categorises four flood types, 16 ML algorithms, 40 data types, 40 evaluation metrics, and 38 common limitations. The results reveal that, despite substantial advancements in the field, critical challenges remain, particularly in terms of data quality, availability, and generalizability of proposed solutions. These gaps underscore the need for more robust, integrative, and contextually adaptive approaches to improve flood prediction systems.
Tipo de publicación: Book Chapter
Publicado en: Advanced Research in Technologies, Information, Innovation and Sustainability
Autores- Brenes, Jose A.
- Castillo, José Daniel Sánchez
Investigadores del CITIC asociados a la publicación
Lic. Jose Daniel Sánchez Castillo
Mag. José Antonio Brenes Carranza
Proyecto asociado a la publicación