Drought forecasting has evolved from empirical and physically based approaches to sophisticated data‐driven methods that integrate climate indices, hydrological models and machine learning.