Key points are not available for this paper at this time.
Climate change has caused an increasing threat of flood disasters, and using artificial intelligence methods to predict floods is now a hot subject in the field of flood prediction. To find out the current situation of the flood prediction research based on artificial intelligence methods, it is essential to summarise the main research focus and direction at present. 612 references on flood forecasting based on AI methods were selected from the Web of Science Core Collection database. The collected articles were analysed visually using CiteSpace and VOSviewer. The results of the study indicate that the overall trend of publications in AI-based flood prediction studies is increasing. In particular, China, the United States and India are the main contributors to research in this research area. The analysis of collaborating institutions shows that Chinese institutions have high activity in this field. The keywords and term analysis show that the research direction of flood prediction based on AI methods mainly focuses on three aspects, which are flood risk assessment, hydrological information prediction and simulation, and integration and improvement of AI algorithms. In recent years, the integration and improvement of artificial intelligence (AI) algorithms have became as a new focal point of research. The incorporation of multiple machine learning or deep learning methods, as well as using additional algorithms to improve the quality of prediction models have received high attention in this flood prediction. In the future, it is important for research efforts to explore these research avenues further, in order to strengthen China's scientific and efficient response capabilities in the face of flood disasters. This study's results can be a reference for researchers to understand the current landscape and emerging frontiers of AI-driven flood prediction research. It will help guide future research directions and strategies and promote the continued development of this field.
Guan et al. (Fri,) studied this question.