Fang et al. (2025) Reference evapotranspiration in Guangxi, China: Spatiotemporal patterns and multi-scale driving mechanisms

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-12-15
• Authors: Rongjie Fang, Linfeng Wang, Junhong Chen, Yue Ben, Shidong Su, Jianying Mo
• DOI: 10.1016/j.ejrh.2025.103038

Research Groups

• Guangxi Key Laboratory of Environmental Pollution Control Theory and Technology, Guilin University of Technology, China
• University Engineering Research Center of Watershed Protection and Green Development, Guangxi, Guilin University of Technology, China
• College of Environmental Science and Engineering, Guilin University of Technology, China
• Naval University of Engineering, Basic Department, China
• Guilin Field Irrigation Experimental Center, China
• Guilin Water and Resources Bureau, China

Short Summary

This study analyzed the spatiotemporal patterns and multi-scale driving mechanisms of reference evapotranspiration (ET0) in Guangxi, China, from 1960 to 2024, revealing an overall declining trend in ET0 primarily driven by meteorological factors, especially sunshine duration, with significant threshold effects.

Objective

• To reveal the spatiotemporal patterns and periodic characteristics of ET0.
• To quantify the independent and synergistic effects of driving factors on ET0 changes across time-frequency domain scales.
• To explain the nonlinear response mechanisms and threshold effects between ET0 and its driving factors.

Study Configuration

• Spatial Scale: Guangxi, China (approximately 237,600 square kilometers), covering latitudes 20°54′N to 26°24′N and longitudes 104°28′E to 112°04′E, utilizing data from 23 meteorological stations.
• Temporal Scale: 1960–2024 (65 years), with analysis of interannual and seasonal variations (Spring: March-May, Summer: June-August, Autumn: September-November, Winter: December-February).

Methodology and Data

• Models used:
  • Penman-Monteith (PM) method (FAO recommended) for ET0 calculation.
  • Theil–Sen Median trend analysis (Sen’s slope estimator).
  • Mann–Kendall (M-K) test for trend significance.
  • Wavelet transform coherence (WTC) and Multiple wavelet coherence (MWC) using the Morlet wavelet function for time-frequency domain analysis.
  • Generalized Additive Models (GAM) with thin-plate regression splines for nonlinear relationships.
  • Piecewise regression for identifying critical threshold points.
  • Monte Carlo methods for WTC and MWC confidence levels.
  • Mann-Whitney U tests for threshold significance.
• Data sources:
  • Meteorological observation data: Daily Surface Climate Dataset of China from 23 meteorological stations in Guangxi. Variables included daily mean, minimum, and maximum air temperatures (Tmean, Tmin, Tmax), mean relative humidity (RH), sunshine duration (SSD), and wind speed (U).
  • Large-scale climate variability (LSCV) indices: Ni˜no 3.4 Index (Ni˜no3.4), Atlantic Oscillation index (AO), Atlantic Multidecadal Oscillation index (AMO), Pacific Decadal Oscillation index (PDO), and North Pacific Oscillation index (NP), obtained from the Earth System Research Laboratory (https://psl.noaa.gov/data/climateindices/list/).

Main Results

• The annual mean ET0 in Guangxi from 1960 to 2024 was 1114.07 mm. An overall declining trend was observed, despite a slight increase after a turning point around 1997, indicating an "evaporation paradox."
• ET0 exhibited significant seasonal differences, peaking in summer (387.49 mm) and reaching its lowest in winter (161.89 mm).
• Spatially, ET0 generally declined across the region, with the most significant decreases in the northwest and the most pronounced fluctuations in the central region.
• Meteorological factors exerted a stronger influence on ET0 variability than large-scale climate variability (LSCV) factors.
• Sunshine duration (SSD) was the most significant individual meteorological driver (average WTC = 0.79, PASC = 75.27%), while the North Pacific Oscillation index (NP) was the most dominant individual LSCV factor (WTC = 0.43, PASC = 21.18%).
• Multi-factor analysis showed that the Tmax-SSD (MWC = 0.91, PASC = 74.01%) and Tmax-U-SSD (MWC = 0.96, PASC = 81.20%) combinations were the strongest meteorological drivers. For LSCV, Ni˜no3.4-NP-PDO (MWC = 0.81, PASC = 27.82%) was the most influential combination.
• The hybrid combination of SSD and the Atlantic Multidecadal Oscillation index (AMO) (MWC = 0.90, PASC = 71.03%) was identified as the most significant dual-factor driver involving both meteorological and LSCV factors.
• Significant nonlinear responses (p < 0.01) and distinct threshold effects were identified for ET0 with Tmax, AMO, and NP. ET0 was significantly higher when Tmax exceeded 24.86 °C, AMO exceeded 0.15, or NP exceeded 1007.49.
• Karst areas exhibited a significantly lower historical mean ET0 (1069.34 mm) compared to non-karst areas (1148.56 mm), with wind dynamics playing a more critical role in ET0 formation mechanisms in karst regions.

Contributions

• Provides a novel, comprehensive ET0 driving paradigm integrating time-frequency domain analysis, nonlinear threshold responses, and cross-system coupling.
• Quantifies the independent and synergistic effects of meteorological and large-scale climate variability factors on ET0 across multiple time-frequency domains, addressing previous research gaps.
• Enhances the mechanistic understanding of ET0 variability and its complex, nonlinear responses to driving factors, including the identification of critical thresholds.
• Offers valuable insights for improving ET0 prediction models and informing regional water resource management, particularly for agricultural irrigation strategies in the context of climate change and specific geomorphological conditions like karst landscapes.

Funding

• Guangxi Key Research and Development Program (Guike-AB24010040, Guike-AB23026045)
• National Natural Science Foundation of China (42561049)
• Guangxi Young Elite Scientist Sponsorship Program (GXYESS2025030)
• Guilin Key Research and Development Program (20210212–2)
• Guangxi Science and Technology Program (Guike-AD25069074)

Citation

@article{Fang2025Reference,
  author = {Fang, Rongjie and Wang, Linfeng and Chen, Junhong and Ben, Yue and Su, Shidong and Mo, Jianying},
  title = {Reference evapotranspiration in Guangxi, China: Spatiotemporal patterns and multi-scale driving mechanisms},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.103038},
  url = {https://doi.org/10.1016/j.ejrh.2025.103038}
}