Water Research in the Age of AI: A Bibliometric Heuristic Analysis for Trends and Opportunities

Water is an essential resource for food, agriculture, health and industry. It is currently facing challenges posed by demographic, economic and climate change. Meanwhile, advances in artificial intelligence and machine learning (AI-ML) offer an opportunity to process large volumes of data to understand complex water-related phenomena better. This study aims to analyse the intersection of Water and AI-ML research, identifying trends shaping current research and giving insights about the future. It uses bibliometric analysis to examine Scopus data, with an incremental, heuristic approach. By the end, the analysis, based on Scival, is also crosschecked with Web of Science (WOS) data. The findings show a great dynamism in this field, with a focus on engineering approaches like "Remote Sensing," "Optical Engineering," and "Photogrammetry." Research topics emphasise detecting, measuring, modelling, and predicting hydrological, geological, and climatic phenomena. There is also a growing cross-disciplinary emphasis on energy, thermodynamics, materials, and chemistry. On the other hand, AI-ML techniques, such as "Artificial Neural Networks", "Random Forest" and "Long Short-Term Memory", are increasingly used. It also seems that the trend is towards studying parameters related to "Water Quality", "Ground Water" and "Water Levels". This particularly benefit the fields of "Agriculture" and "Water Resources Management". At the level of international collaboration, an accentuated concentration is noted with a small number of prolific countries (China, USA, India). However, industrial collaboration remains relatively weak, presenting both opportunities and challenges. Finally, this research seems to be taking a direction that is still in its infancy, due to its precocity. Nevertheless, it presents considerable potential for development, particularly in terms of international and industrial collaboration. As a perspective, the expert information comparison would certainly be beneficial to enrich this bibliometric analysis and improve the robustness of the conclusions.