Lu et al. (2025) Vegetation evolution and water sensitivity analysis in the source region of the Yangtze River and the Yellow River under the combined drive of energy-temperature-water

Identification

• Journal: Ecological Indicators
• Year: 2025
• Date: 2025-11-20
• Authors: Jie Lu, Tianling Qin, Xizhi Lv, Denghua Yan, Shuhua Xu, Zhe Yuan, Jianwei Wang, Weizhi Li, Haoyue Gao
• DOI: 10.1016/j.ecolind.2025.114428

Research Groups

• State Key Laboratory of Water Cycle and Water Security, China Institute of Water Resources and Hydropower Research, Beijing, China
• Henan Key Laboratory of Yellow Basin Ecological Protection and Restoration, Yellow River Institute of Hydraulic Research, Zhengzhou, China
• Changjiang River Scientific Research Institute of Changjiang Water Resources Commission, Wuhan, China

Short Summary

This study analyzed the temporal and spatial evolution of vegetation in the source region of the Yangtze and Yellow Rivers, revealing that 82.26% of the regional vegetation significantly improved from 1982 to 2020. It found that temperature and energy primarily influenced vegetation indirectly through water sources, while precipitation and shallow soil water were the most significant direct water source factors.

Objective

• To reveal the temporal and spatial evolution of vegetation in the source region of the Yangtze River and the Yellow River.
• To assess the multi-scale coupling oscillation of vegetation and various factors in the time–frequency domain, screen key driving factors, and construct a causal chain analysis technique of regional 'energy distribution-temperature change-water source composition-vegetation growth' to reveal the driving mechanism of vegetation change.
• To identify sensitive water sources affecting vegetation growth by zoning (different geomorphic units) and classification (different vegetation types).

Study Configuration

• Spatial Scale: Source region of the Yangtze River and the Yellow River, located in the eastern margin of the Qinghai-Tibet Plateau, covering approximately 2.644 × 10^5 square kilometers. Data resolutions ranged from 30 meters to 0.25 degrees, with most data resampled to 250 meters × 250 meters for causal analysis.
• Temporal Scale: 1982 to 2020 (39 years), with data primarily at a monthly temporal resolution.

Methodology and Data

• Models used:
  • Sen-Mann-Kendall trend analysis
  • Wavelet Coherence (WTC) and Multi-Wavelet Coherence (MWC)
  • Structural Equation Model (SEM), specifically Partial Least Squares Structural Equation Model (PLS-SEM)
  • PER-Kriging interpolation method
  • Spatial–temporal adaptive reflectance fusion model (ESTARFM)
  • Maximum Value Synthesis (MVC) method
• Data sources:
  • Satellite/Reanalysis:
    • Global Land Data Assimilation System (GLDAS) Noah Land Surface Model (versions 2.0 and 2.1) for energy factors (net radiation, latent heat net flux, sensible heat net flux, heat flux), surface temperature, and water source factors (storm surface runoff, snowmelt water, soil water at 4 depths, groundwater storage).
    • MOD13Q1 Normalized Difference Vegetation Index (NDVI) (2000-2020, 250 m resolution).
    • GIMMS NDVI3g (1981-2006, 8 km resolution).
    • ASTER Global Digital Elevation Model (GDEM) V3 (30 m resolution).
  • Observation/Other:
    • National Meteorological Information Center of China Meteorological Administration (daily average temperature and daily precipitation from 23 meteorological stations, 1982-2020).
    • Resource and Environmental Science Data Center of the Chinese Academy of Sciences (1:1,000,000 landform type map, China 30 m land use data set for 1980-2020).

Main Results

• From 1982 to 2020, the annual average NDVI in the source region fluctuated between 0.247 and 0.287, showing an 'increase–decrease-increase' evolution. Vegetation coverage significantly improved, with 82.26% of the regional vegetation showing improvement, particularly after 2006.
• Spatially, vegetation improvement was higher in the southeast and lower in the northwest.
• Multi-wavelet coherence (AWC) values between NDVI and energy, water sources, and temperature reached 0.96, 0.99, and 0.93, respectively, indicating strong synergistic effects.
• Structural Equation Model (SEM) results showed that the total effects of temperature, energy, and water source factors on vegetation were 0.96, 0.78, and 0.34, respectively.
• Energy and temperature primarily affected vegetation indirectly through water sources, with indirect influence coefficients of 0.96 for energy and 0.29 for temperature.
• Water source factors directly affected vegetation, with precipitation (0.91) and shallow soil water (0.86) having the most significant effects. Surface runoff (0.47) and snowmelt water (0.39) also showed notable impacts.
• The sensitive water sources for vegetation growth were identified as precipitation and surface soil water.
• Spatially, precipitation and soil water were crucial for the water supply of cultivated land, forest land, grassland, and wetland vegetation. Surface runoff (0.23) and snowmelt water (0.13) had a significant impact on grassland.
• The influence intensity sequence of water source factors on vegetation across main geomorphic types was consistently: precipitation > soil water > surface runoff > snowmelt water > groundwater storage.

Contributions

• Systematically revealed the temporal and spatial evolution of vegetation in the source region of the Yangtze and Yellow Rivers over a long period (1982-2020).
• Developed and applied a novel causal chain analysis technique ('energy distribution-temperature change-water source composition-vegetation growth') using multi-cross wavelet analysis and structural equation modeling to uncover the complex, multi-factor driving mechanisms of vegetation change.
• Quantified the direct and indirect effects of energy, temperature, and water sources on vegetation, highlighting the dominant indirect role of energy and temperature through water sources, which was previously vague in literature.
• Identified sensitive water sources for different geomorphic units and vegetation types, providing spatially explicit and targeted insights for vegetation restoration and ecological management in alpine plateau areas.
• Overcame limitations of traditional statistical methods by capturing multi-factor synergy and nonlinear responses, and improved interpretability compared to 'black box' machine learning models.

Funding

• National Key Research and Development Program of China (Grant No. 2022YFC3201705)
• National Science Fund Project (Grant No. U2443205)
• National Science Fund Project (Grant No. 52130907)
• Special Project on Basic Scientific Research Funds of China Institute of Water Resources and Hydropower Research (JZ110145B0032025)
• Five Major Excellent Talent Programs of IWHR (WR0199A012021)

Citation

@article{Lu2025Vegetation,
  author = {Lu, Jie and Qin, Tianling and Lv, Xizhi and Yan, Denghua and Xu, Shuhua and Yuan, Zhe and Wang, Jianwei and Li, Weizhi and Gao, Haoyue},
  title = {Vegetation evolution and water sensitivity analysis in the source region of the Yangtze River and the Yellow River under the combined drive of energy-temperature-water},
  journal = {Ecological Indicators},
  year = {2025},
  doi = {10.1016/j.ecolind.2025.114428},
  url = {https://doi.org/10.1016/j.ecolind.2025.114428}
}