Wang et al. (2025) Dominant role of climatic water availability in net ecosystem productivity in China's drylands: A comparison with atmospheric water demand and soil moisture

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-11-01
• Authors: Zhen Wang, Mingyang Jin
• DOI: 10.1016/j.ejrh.2025.102891

Research Groups

• School of Surveying and Land Information Engineering, Henan Polytechnic University, Jiaozuo, Henan, China

Short Summary

This study investigated the dominant water stress factors (climatic water availability, atmospheric water demand, and soil moisture) influencing net ecosystem productivity (NEP) in China's drylands over 40 years. It found that climatic water availability (SPEI) is the primary driver for 44.04% of natural vegetation, though its dominance shifts to atmospheric water demand (VPD) and soil moisture (SM) as aridity decreases, with specific aridity index thresholds.

Objective

• To identify the dominant water stress factors (climatic water availability represented by SPEI, atmospheric water demand by VPD, and soil moisture by SM) influencing the carbon sequestration capacity (Net Ecosystem Productivity, NEP) of natural vegetation in China's drylands.
• To assess how the relative effects of these factors vary with the regional aridity index (AI) and to identify critical AI thresholds where the primary influencing factor shifts.

Study Configuration

• Spatial Scale: China's drylands, covering approximately 6.6 million square kilometers, focusing on natural vegetation ecosystems (46.62% of dryland area), including grasslands (78.77%), shrublands (7.42%), and forests (13.81%).
• Temporal Scale: Nearly 40 years, specifically from 1981/1982 to 2019.

Methodology and Data

• Models used:
  • Boreal Ecosystem Productivity Simulator (BEPS) for generating NEP data.
  • Penman-Monteith method for potential evapotranspiration (PET) calculation.
  • "SPEI" R package for Standardized Precipitation Evapotranspiration Index (SPEI-12) calculation.
  • Linear least squares regression, piecewise linear regression, Theil-Sen estimator, and Mann-Kendall test for trend analysis.
  • Pearson correlation analysis for coupling relationships.
  • Variance Inflation Factor (VIF) for multicollinearity assessment.
  • Ridge regression model and Partial Least Squares (PLS) regression for analyzing relative effects, with 20-fold cross-validation for parameter optimization.
• Data sources:
  • Global NEP simulation data (1981-2019): National Ecological Science Data Center (daily, 0.072727° × 0.072727° spatial resolution).
  • Meteorological and soil moisture data (1982-2019): ERA5-Land reanalysis (monthly, 0.1° spatial resolution), including precipitation, potential evapotranspiration, soil moisture (0-100 cm), air temperature, and dew point temperature.
  • Vapor Pressure Deficit (VPD) calculation: Method recommended by FAO Irrigation and Drainage Paper No. 56.
  • Eddy-covariance observations: FLUXNET 2015 Tier 2 (39 dryland natural vegetation sites) for validation.
  • Auxiliary datasets: Aridity Index (AI) data (ratio of annual precipitation to potential evapotranspiration) and MODIS MCD12C1 land cover maps (0.05° original spatial resolution).
  • All gridded datasets were resampled to a 0.1° spatial resolution using bilinear interpolation.

Main Results

• The carbon sequestration capacity (NEP) of natural vegetation in China’s drylands increased at an average rate of 0.49 ± 0.13 g C m⁻² per year (equivalent to 4.9 ± 1.3 g C m⁻² per decade) from 1982 to 2019.
• Water stress progressively intensified: SPEI and soil moisture (SM) showed significant downward trends from 1991 to 2019 (slopes of -0.068 ± 0.018 and -0.091 ± 0.015, respectively), while vapor pressure deficit (VPD) showed a significant upward trend (slope of 0.076 ± 0.017) during the same period.
• Climatic water availability (SPEI) was the dominant water stress factor for 44.04% of the natural vegetation in China’s drylands. Atmospheric water demand (VPD) dominated 34.74%, and soil moisture (SM) dominated 21.22%.
• As the aridity index (AI) increases (i.e., climate becomes wetter), the relative effects of VPD and SM gradually surpass that of SPEI.
• Specific AI thresholds were identified: VPD's relative effect surpassed SPEI when AI exceeded 0.36. SM's relative effect surpassed SPEI when AI exceeded 0.91 (and was nearly equivalent between AI 0.41 and 0.91).
• The negative response of NEP to VPD was widespread, with 86.01% of vegetated areas showing a negative impact on NEP due to rising VPD.

Contributions

• This study provides the first comprehensive comparison of the relative effects of climatic water availability (SPEI), soil moisture (SM), and atmospheric water demand (VPD) on Net Ecosystem Productivity (NEP) in China's drylands.
• It identifies SPEI as the overall dominant water stress factor for carbon sequestration in China's dryland natural vegetation, a finding validated by both regression models and flux tower observations.
• The research quantifies the dynamic shift in dominant water stress factors along an aridity gradient, establishing specific Aridity Index (AI) thresholds (0.36 for VPD, 0.91 for SM) where their influence on NEP surpasses that of SPEI.
• The findings highlight that declining climatic water availability is the foremost water stress constraining the carbon sequestration capacity of dryland vegetation ecosystems in China, offering crucial insights for climate change impact assessments and management strategies.

Funding

• The study acknowledges support from the National Ecological Science Data Center for providing NEP data, ERA5-Land for meteorological data, and Henan Polytechnic University for equipment and laboratory facilities. No specific project or program reference codes were provided.

Citation

@article{Wang2025Dominant,
  author = {Wang, Zhen and Jin, Mingyang},
  title = {Dominant role of climatic water availability in net ecosystem productivity in China's drylands: A comparison with atmospheric water demand and soil moisture},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102891},
  url = {https://doi.org/10.1016/j.ejrh.2025.102891}
}