Mapping a Quarter-Century of Fuzzy-Based Models for Groundwater Level Estimation: A Bibliometric Analysis

چکیده مقاله

Groundwater, the planet’s largest freshwater reservoir, faces mounting stress from overuse and climate change, demanding interpretable, uncertainty-aware modeling tools. Fuzzybased modeling is offered as a unique solution, yet its global research landscape, scientific evolution, intellectual structure, and global diffusion remain unmapped. To address this gap, we present the first focused bibliometric synthesis of fuzzy-based approaches for groundwater level estimation, analyzing 189 Web of Science–indexed articles published between 2000 and 2025 using Bibliometrix and VOSviewer. Our analysis quantifies publication growth, identifies leading countries, institutions, and authors, and maps thematic clusters. Results reveal a 19.8% annual growth rate since 2016, with Iran dominating in publications and the United States achieving the highest citations. Key institutions include Islamic Azad University, University of Tabriz, and University of Tehran, while leading authors such as Kisi, El-Shafie, and Nourani anchor the intellectual structure of the field. Science-mapping demonstrates the dominance of hybrid frameworks, particularly those integrating wavelet transforms, metaheuristics, or adaptive neuro-fuzzy inference systems (ANFIS). Keyword co-occurrence networks (minimum occurrence = 5) and temporal overlay visualizations demonstrate a paradigm shift from standalone fuzzy systems toward climate-aware, physicsinformed hybrids. While fuzzy models offer high interpretability and robustness in handling uncertainty, challenges remain regarding spatial scalability and integration with large-scale physical models. This study consolidates a scattered body of knowledge and establishes fuzzy-based groundwater modeling as a maturing sub-discipline. While these models offer distinct strengths such as interpretability and the ability to handle epistemic uncertainty, they also present weaknesses, including challenges in spatial scalability and data dependency. We identify critical frontiers, including explainable AI integration, improved spatial scalability, and applications in data-scarce transboundary aquifers. Future research should prioritize cross-paradigmatic collaboration to bridge data-driven and process-based modeling, enhancing both scientific rigor and policy relevance, particularly in the context of escalating water insecurity driven by climate change