Flood Forecasting Using Artificial Intelligence (AI): A Prisma – Guided Systematic Review
Abstract
Subject and purpose of work
The purpose of the study is to examine the implementation of artificial intelligence (AI) techniques to forecast floods.
Materials and methods
This study systematically reviews 42 papers published between 1989 and 2025. The dataset was extracted from the Scopus and Web of Science databases using predefined keywords and inclusion-exclusion criteria. Following the PRISMA-2020 framework, the analysis investigates AI techniques, statistical methods, and commonly used keywords in flood forecasting.
Results
The s tudy r eveals t hat m achine l earning ( ML) a nd d eep l earning ( DL) a re w idely applied AI techniques, with the root mean square error (RMSE) that has been employed for flood forecasting. The keywords, such as “artificial intelligence,” “prediction,” “flood forecasting,” and “floods,” are frequently used in this domain.
Conclusion
The s tudy concludes t hat advanced A I techniques enhance forecasting accuracy, improve early warning systems, and promote the development of evidence-based disaster reduction strategies.
© 2026 Somanath Kumar Mishra, Prasanta Patri, published by John Paul II University in Biała Podlaska
This work is licensed under the Creative Commons Attribution 4.0 License.