Zhang et al. (2026) Divergent impacts of flash drought drivers on alpine ecosystem resilience

Identification

• Journal: Journal of Environmental Management
• Year: 2026
• Date: 2026-03-26
• Authors: Yu Zhang, Yixue Hong, Xiao Yang, Xin Wang, Xiaojun Zhang, Jinling Pei, Zhaohui Luo, Lixuan Zhou, Xijun Yu Xijun Yu, Huhu Kang, Qiangqiang Lu
• DOI: 10.1016/j.jenvman.2026.129445

Research Groups

• Ministry of Ecology and Environment Key Laboratory of Urban Ecological Environment Simulation and Protection, South China Institute of Environmental Sciences, Ministry of Ecology and Environment of China, Guangzhou, 510530, China
• State Key Laboratory of Biocontrol, School of Ecology, Sun Yat-Sen University, Shenzhen, 518107, China
• School of Geography and Tourism, Shaanxi Normal University, Xi'an, 710119, China
• Shaanxi Key Laboratory of Qinling Ecological Security, Xi'an Botanical Garden of Shaanxi Province (Institute of Botany of Shaanxi Province), Xi'an, 710061, China

Short Summary

This study reveals divergent impacts of flash drought drivers on alpine ecosystem resilience in the Yarlung Tsangpo River Basin, finding that while soil moisture deficits consistently weaken resilience, temperature-driven events can temporarily enhance it due to meltwater subsidies, masking underlying vulnerabilities.

Objective

• To quantify the contribution weights of different drivers (temperature, precipitation, soil moisture) to flash drought occurrence and map their spatial patterns.
• To analyze the variations in flash drought frequency, intensity, duration, and recovery time along gradients of driver contribution weights.
• To evaluate the spatial differences in the evolution of ecosystem resilience in response to flash droughts with distinct driving mechanisms.

Study Configuration

• Spatial Scale: Yarlung Tsangpo River Basin (YTRB), Southern Tibetan Plateau, China.
• Temporal Scale: 2000–2023.

Methodology and Data

• Models used:
  • Principal Component Analysis Regression Flash Drought model (PRFD) for multivariate drought identification.
  • Kalman filter-based Bayesian linear dynamic model to quantify ecosystem resilience using first-order lagged temporal autocorrelation (TAC) of kernel Normalized Difference Vegetation Index (kNDVI).
  • Mann-Kendall (MK) test for trend detection.
  • changepoint.np package for change point detection.
  • Geographically Weighted Regression (GWR) for spatial statistical analysis.
• Data sources:
  • ERA5-Land hourly reanalysis products (0.1° resolution): precipitation, air temperature, soil moisture (0-7 cm depth), dew point temperature, relative humidity, snow water equivalent, surface downward radiation.
  • Ground measurements: daily soil moisture, temperature, and precipitation from meteorological stations in China.
  • AVH13C1 dataset of Long Term Data Record (LTDR): daily Normalized Difference Vegetation Index (NDVI) (0.05° spatial resolution).
  • MOD13A3 monthly NDVI product from Moderate Resolution Imaging Spectroradiometer (MODIS) (1 km spatial resolution).
  • First Landsat-derived annual land cover product of China (CLCD).

Main Results

• Flash droughts in the Yarlung Tsangpo River Basin (YTRB) intensified from 2000 to 2023, with frequency increasing at 0.56 events per year and duration at 2.25 pentads per year.
• Flash drought drivers exhibited spatial heterogeneity, with events nearly equally driven by temperature (47.9% of basin area) and soil moisture (49.9%).
• Widespread ecosystem resilience degradation was observed across 92.9% of the basin, with a significant regime shift detected in 2010, after which resilience remained persistently low.
• A divergent impact of flash drought drivers was identified:
  • Soil moisture deficits-driven flash droughts consistently weakened ecosystem resilience.
  • Temperature-driven flash droughts triggered a transient resilience enhancement, with resilience strengthening proportionally to temperature's contribution (R² = 0.78 for mean TAC, R² = 0.64 for TAC trend). This is attributed to meltwater subsidies from thawing permafrost and physiological thermal acclimation.
• Ecosystem type significantly modulated resilience: forests demonstrated remarkable buffering capacity and the lowest degradation rate (0.00039 per year), while grasslands and shrublands in the mid-upper reaches suffered severe resilience loss (grasslands: 0.00064 per year).
• Increasing flash drought frequency, intensity, and duration were strongly associated with declining resilience, with high-frequency events being particularly detrimental.

Contributions

• Provides the first comprehensive assessment of divergent impacts of thermodynamic and hydraulic flash drought drivers on alpine ecosystem resilience, challenging the assumption of uniform drought impacts.
• Identifies a novel "masked resilience" phenomenon in alpine ecosystems, where temperature-driven flash droughts temporarily enhance resilience due to meltwater subsidies, potentially obscuring long-term vulnerabilities.
• Quantifies the distinct resilience mechanisms and degradation rates across different alpine ecosystem types (e.g., forests, grasslands, shrublands), highlighting the critical role of forests as "resilience anchors."
• Offers a scientific basis for developing targeted, differentiated adaptive management strategies for vulnerable high-altitude environments, emphasizing the protection of high-altitude forests and dynamic adaptation for grasslands.

Funding

• Guangdong Basic and Applied Basic Research Foundation (Grants: 2024A1515510007)
• National Natural Science Foundation of China (Grants: 42307584, 42301053)

Citation

@article{Zhang2026Divergent,
  author = {Zhang, Yu and Hong, Yixue and Yang, Xiao and Wang, Xin and Zhang, Xiaojun and Pei, Jinling and Luo, Zhaohui and Zhou, Lixuan and Yu, Xijun Yu Xijun and Kang, Huhu and Lu, Qiangqiang},
  title = {Divergent impacts of flash drought drivers on alpine ecosystem resilience},
  journal = {Journal of Environmental Management},
  year = {2026},
  doi = {10.1016/j.jenvman.2026.129445},
  url = {https://doi.org/10.1016/j.jenvman.2026.129445}
}