Wu et al. (2026) Vegetation productivity responses and environmental controls under flash drought stress in the Yellow River Basin
Identification
- Journal: Journal of Hydrology Regional Studies
- Year: 2026
- Date: 2026-11-16
- Authors: Jincheng Wu, Kunxia Yu, Peng Li, Zhanbin Li, Shengde Yu, Lu Jia, Xue Li, Yue ZiHui, Zhi Yang
- DOI: 10.1016/j.ejrh.2026.104232
Research Groups
- State Key Laboratory of Water Engineering Ecology and Environment in Arid Area, Xi’an University of Technology, Xi’an, Shaanxi, China
- Ecohydrology Research Group, Department of Earth and Environmental Sciences, University of Waterloo, Waterloo, Ontario, Canada
- Ningxia Water Resources Research Institute, Yinchuan, Ningxia, China
- Xinjiang Uygur Autonomous Region Ecological Water Conservancy Research Center, Urumqi, Xinjiang, China
Short Summary
This study investigated the spatiotemporal patterns of flash droughts and their environmental controls on vegetation productivity in the Yellow River Basin from 2000-2020, finding that soil moisture primarily limits gross primary productivity (GPP) in the middle-lower reaches while vapor pressure deficit (VPD) dominates in upstream regions.
Objective
- Quantify the coupling relationships among GPP, soil moisture (SM), and VPD, and disentangle the independent effects of soil and atmospheric water stress on vegetation productivity.
- Identify flash drought events in the Yellow River Basin and characterize their spatiotemporal evolution in terms of frequency and duration.
- Reveal the nonlinear threshold responses of GPP to variations in water availability, characterize shifts in ecosystem water limitation regimes along climatic gradients and among different vegetation types, and further identify the dominant controls of key meteorological drivers.
Study Configuration
- Spatial Scale: Yellow River Basin (YRB), China, covering approximately 795,000 km².
- Temporal Scale: Two decades (2000–2020), with data analyzed at annual, 8-day, and 5-day (pentad) resolutions.
Methodology and Data
- Models used: Partial correlation analysis, percentile-based flash drought identification algorithm (using Gringorten plotting position formula), SM–VPD decoupling framework, Random Forest regression model, SHAP (SHapley Additive exPlanations) values.
- Data sources:
- Gross Primary Productivity (GPP): Terra MOD17A2HGF Version 6.1 (8-day, 500 m spatial resolution), GOSIF-GPP Version 2 (8-day, 0.05° spatial resolution) for robustness checks.
- Soil Moisture (SM) and Potential Evaporation (PET): ERA5-Land reanalysis dataset (daily, 0.1° × 0.1° spatial resolution), with SM calculated for the upper 1 m (0–7 cm, 7–28 cm, 28–100 cm layers).
- Meteorological variables (air temperature (TEMP), relative humidity (RHU), wind speed (WIN), sunshine duration (SSD), precipitation (PRE)): CN05.1 gridded dataset (daily, 0.25° × 0.25° spatial resolution).
- Land-cover types: Reclassified International Geosphere–Biosphere Programme (IGBP) scheme.
- Aridity Index (AI) for climatic zone classification.
Main Results
- Flash droughts exhibited significant spatial heterogeneity, with the highest frequency and duration concentrated in the middle and lower reaches of the Yellow River Basin, particularly the southern Loess Plateau and Guanzhong Plain.
- Localized hotspots experienced up to 24 flash drought events and a cumulative duration of 1280 days (256 pentads).
- In these high-frequency regions, rapid soil moisture depletion induced persistent vegetation water stress, with soil moisture (SM) identified as the dominant limiting factor for GPP.
- Conversely, upstream plateau regions experienced fewer flash droughts, where vapor pressure deficit (VPD) served as the primary limiting factor for ecosystem productivity.
- GPP showed nonlinear responses to SM and VPD, peaking under moderate water conditions (50th-60th percentiles for both SM and VPD) and declining sharply under severe soil dryness or atmospheric water demand.
- Cropland GPP demonstrated a strong dependence on SM supply, while forests exhibited greater sensitivity to VPD and enhanced resilience to climatic fluctuations.
- Sunshine duration (SSD) and VPD were identified as the largest statistical contributors to GPP variability during flash droughts, based on Random Forest–SHAP analysis. Optimal GPP was predicted around 26712 seconds (7.42 hours) of SSD, 687.75 Pa of VPD, and 0.33 m³⋅m⁻³ of SM.
Contributions
- Provided a comprehensive analysis of short-term nonlinear responses and independent contributions of soil and atmospheric water stress on GPP during flash droughts, addressing a gap in studies primarily focused on seasonal or interannual drought conditions.
- Elucidated the dynamic mechanisms linking flash drought evolution to vegetation productivity responses, moving beyond mere identification and climatological characteristics.
- Quantified the nonlinear responses of vegetation to compound soil-atmospheric drought stress and characterized spatial shifts in dominant water constraints across climatic gradients and vegetation types.
- Integrated multi-source datasets and advanced modeling techniques (pentad-scale flash drought identification, multi-product SM–VPD decoupling analysis, and Random Forest–SHAP modeling) to offer new hydrological insights for the Yellow River Basin.
Funding
- National Natural Science Foundation of China (52579021)
- Major Science and Technology Special Project of Xinjiang Uygur Autonomous Region (2025A08013–3)
- Key Research and Development Program of Shaanxi Province (2023-ZDLSF-60)
Citation
@article{Wu2026Vegetation,
author = {Wu, Jincheng and Yu, Kunxia and Li, Peng and Li, Zhanbin and Yu, Shengde and Jia, Lu and Li, Xue and ZiHui, Yue and Yang, Zhi and Xu, Yan},
title = {Vegetation productivity responses and environmental controls under flash drought stress in the Yellow River Basin},
journal = {Journal of Hydrology Regional Studies},
year = {2026},
doi = {10.1016/j.ejrh.2026.104232},
url = {https://doi.org/10.1016/j.ejrh.2026.104232}
}
Original Source: https://doi.org/10.1016/j.ejrh.2026.104232