Water balance–based evaluation and attribution of multi-source evapotranspiration datasets in the Qinling–Daba Mountains, China

Study region: The Qinling–Daba Mountains, China. Study focus: Due to the complex topography and heterogeneous climatic conditions of the Qinling–Daba (Qinba) Mountains, substantial uncertainties remain among evapotranspiration (ET) datasets with varying spatial and temporal resolutions. In this study, we evaluated 18 ET datasets using the water balance and Triple Collocation, investigated the factors influencing their accuracy through Spearman's rank correlation analysis across the Qinba Mountains. New hydrogeological insights from the region: In the Qinba Mountains, most ET datasets exhibit a distinct spatial pattern of lower values in the northwest and higher values in the southeast, with annual ET ranging from 400 to 850 mm·yr⁻¹ and clustering around 600 mm·yr⁻¹, accompanied by pronounced seasonal and interannual variability. Among the evaluated datasets, GLEAM4 demonstrates the highest overall performance at both monthly and seasonal scales, followed by PMLV2, GLASS, and ETCR, with all datasets performing better during wet seasons than dry seasons. From 2007–2017, most datasets show strong consistency in interannual variations, particularly a common fluctuation during 2013–2014, indicating high reliability in capturing regional ET dynamics. Topographic complexity is identified as the dominant factor limiting ET model accuracy, whereas climatic energy conditions exert a positive influence on model performance. Overall, this study provides a reference for the selection of ET datasets in the Qinba Mountains and other mountainous regions characterized by complex terrain and climatic heterogeneity.