Research Progress on Estimation Methods of Forest Evapotranspiration Based on Bibliometrics

Forest evapotranspiration (ET), a core process of water vapor exchange between forest ecosystems and the atmosphere, is crucial for global carbon and water cycles and ecosystem stability. However, its high-precision estimation faces challenges arising from complex forest structures and multi-factor driving mechanisms. Based on bibliometrics, this study visually analyzed 1,427 relevant papers from the Web of Science Core Collection (2005-2025) to summarize research status, hotspots and frontiers. Results show continuous growth in publications over two decades, peaking during 2017-2022. Journal co-occurrence reveals that Agricultural and Forest Meteorology ranks first, contributing 212 papers and 10,438 total citations with an average of 49.24 cites per article. The Chinese Academy of Sciences, University of CAS and USDA form a close collaboration network led by 279 core authors. Hotspots concentrate on eddy covariance (460 occurrences), remote sensing inversion (134) and machine learning (124, rapidly rising since 2017). Eddy covariance remains the “gold standard”; remote sensing breaks spatiotemporal limits by integrating multi-source data; machine learning, exhibiting the greatest advances, improves accuracy by 45% in complex environments (burst intensity 17.91 since 2017), promoting hybrid “physical mechanism + data-driven” models. Research evolved through three stages: traditional observation dominance (2005-2010), physical model optimization (2010-2016), and intelligent algorithm innovation (2017-present), with applications spanning ecological assessment and water resource management.