Responses of stable vegetation to drought accumulation timescales and time lags in the Weihe River Basin
- 1School of Resources and Environmental Engineering, Tianshui Normal University , , ,
Abstract
Introduction: This study aimed to distinguish drought accumulation timescale from response timing in stable vegetation in the Weihe River Basin.
Methods: We analyzed MODIS enhanced vegetation index (EVI) and ERA5-Drought standardized precipitation evapotranspiration index (SPEI) data for 2003–2024 on the native 0.25° drought grid. Persistent land-cover classification required at least 18 valid annual classifications and a modal-class fraction of at least 80%. Four accumulation timescales (1, 3, 6, and 12 months) and 13 window-end lags (0–12 months) formed 52 response–exposure combinations. We controlled for growing-season month effects and a linear trend. We applied maximum-statistic tests with synchronized 2-year moving-block resampling and controlled the false discovery rate across cells using the Benjamini–Hochberg procedure.
Results: The area-weighted mean of grid-cell partial correlations peaked at SPEI-12, lag 0 (r = 0.3216), with a mean slope of 0.00971 EVI units per SPEI unit. Of 190 eligible cells, 126 showed a corrected positive association, covering 69.68% of eligible stable-vegetation area. Grasslands and high-elevation areas had stronger descriptive responses. Relaxing valid-area coverage from 0.70 to 0.50 retained the basin-wide peak. Across detected cells, the median bootstrap reselection frequency of the observed best timescale–lag combination was 53.1%. Across 2,252 SPEI-3 grid-cell drought events, the mean EVI residual reached −0.00909 one month after onset (95% onset-year block-bootstrap interval, −0.01448 to −0.00439). Among 2,105 uncensored events, meteorological drought termination was confirmed a median of 3 months after the last dry month (interquartile range, 2–3 months).
Discussion: The dominant association links vegetation condition to the 12-month water balance ending in the response month. These findings support monitoring that combines long-term water balance, early vegetation responses, and continued vegetation observations after meteorological drought termination.
Related
- Earlier green-up exacerbates northern vegetation sensitivity to flash drought
- Similar evapotranspiration responses mask contrasting modes of water access during drought in California
- Two clones, two paths: divergent intrinsic xylem traits and hydraulic responses to prolonged drought in poplar
- Should I stay or should I go? Frog breeding responses to drought events and alien predators
- Drought responses of a Norway spruce forest on drained peat soil: combining sap-flow sensors, eddy-covariance, meteorological, soil and UAV data
BibTeX
@article{hu2026responses,
title = {{Responses of stable vegetation to drought accumulation timescales and time lags in the Weihe River Basin}},
author = {Xiangxiang Hu and Yaya Shi},
journal = {Frontiers in plant science},
year = {2026},
month = sep,
doi = {10.3389/fpls.2026.1974112},
url = {https://europepmc.org/article/PMC/PMC13661968},
note = {CC BY},
}
RIS
TY - JOUR
TI - Responses of stable vegetation to drought accumulation timescales and time lags in the Weihe River Basin
AU - Xiangxiang Hu
AU - Yaya Shi
PY - 2026
DA - 2026/09/30
JO - Frontiers in plant science
DO - 10.3389/fpls.2026.1974112
UR - https://europepmc.org/article/PMC/PMC13661968
AB - Introduction: This study aimed to distinguish drought accumulation timescale from response timing in stable vegetation in the Weihe River Basin. Methods: We analyzed MODIS enhanced vegetation index (EVI) and ERA5-Drought standardized precipitation evapotranspiration index (SPEI) data for 2003–2024 on the native 0.25° drought grid. Persistent land-cover classification required at least 18 valid annual classifications and a modal-class fraction of at least 80%. Four accumulation timescales (1, 3, 6, and 12 months) and 13 window-end lags (0–12 months) formed 52 response–exposure combinations. We controlled for growing-season month effects and a linear trend. We applied maximum-statistic tests with synchronized 2-year moving-block resampling and controlled the false discovery rate across cells using the Benjamini–Hochberg procedure. Results: The area-weighted mean of grid-cell partial correlations peaked at SPEI-12, lag 0 (r = 0.3216), with a mean slope of 0.00971 EVI units per SPEI unit. Of 190 eligible cells, 126 showed a corrected positive association, covering 69.68% of eligible stable-vegetation area. Grasslands and high-elevation areas had stronger descriptive responses. Relaxing valid-area coverage from 0.70 to 0.50 retained the basin-wide peak. Across detected cells, the median bootstrap reselection frequency of the observed best timescale–lag combination was 53.1%. Across 2,252 SPEI-3 grid-cell drought events, the mean EVI residual reached −0.00909 one month after onset (95% onset-year block-bootstrap interval, −0.01448 to −0.00439). Among 2,105 uncensored events, meteorological drought termination was confirmed a median of 3 months after the last dry month (interquartile range, 2–3 months). Discussion: The dominant association links vegetation condition to the 12-month water balance ending in the response month. These findings support monitoring that combines long-term water balance, early vegetation responses, and continued vegetation observations after meteorological drought termination.
SN - 1664-462X
ER -