Guido Ascenso, Matteo Giuliani, Jorge Pérez‐Aracil, Sancho Salcedo‐Sanz, Claudia Bertini, Paolo Bonetti, Gustau Camps‐Valls, Miguel Ángel Fernández Torres, Andrea Ficchì, Nora Linscheid, Martina Merlo, Giulio Palcic, Markus Reichstein, Marcello Restelli, Andrea Toreti, Eliot Walt, Andrea Castelletti
Abstract Machine learning (ML) has emerged as a key tool in drought research, with applications growing rapidly over the past 20 years. While several reviews have described specific ML methods and their use in forecasting and monitoring, a comprehensive assessment of trends, gaps, and emerging challenges is lacking. Here, we analyze two decades of literature to map the evolution of ML in drought science. We find exponential growth since 2013, driven largely by forecasting and monitoring studies, while impact assessment and explainable artificial intelligence (XAI) remain underexplored. Geographic analysis highlights significant gaps in drought‐prone regions such as Africa and South America. The field shows slow adoption of advanced ML architectures and limited use of large data sets, coupled with reproducibility challenges due to restricted code and data sharing. Addressing these issues is critical to advance ML‐based drought risk management and climate adaptation.