Yanqi et al. (2026) Agricultural flash drought risk assessment under climate change based on crop growth stages

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-10-31
• Authors: Wei Yanqi, Zhou Yuliang, Jin Juliang, Haichao Li, Ming bo, WANG Zongzhi, Yi Cui, TAO Jin, Ping Zhou
• DOI: 10.1016/j.ejrh.2026.104158

Research Groups

• College of Civil Engineering, Joint Laboratory of Artificial Intelligence for Water Security, Hefei University of Technology, Hefei, China
• School of Energy and Power Engineering, Xihua University, Chengdu, China
• School of Geography and Environment, Mianyang Normal University, Mianyang, China
• State Key Laboratory of Hydroscience and Engineering, Department of Hydraulic Engineering, Tsinghua University, Beijing, China
• Department of Hydrology and Water Resources, State Key Laboratory of Eco-hydraulics in Northwest Arid Region, Xi'an University of Technology, Xi'an, China
• State Key Laboratory of Water Disaster Prevention, Nanjing Hydraulic Research Institute, Nanjing, China

Short Summary

This study developed a crop-growth-stage-based framework to assess agricultural flash drought (AFD) risk under climate change, revealing that future AFD events, particularly during reproductive stages, will lead to increased maize yield losses despite potentially shorter durations, primarily driven by rising maximum temperatures.

Objective

• To identify agricultural flash drought (AFD) processes under historical and future climate scenarios and analyze their temporal evolution in terms of duration and severity.
• To estimate the occurrence frequency of AFD and quantify the associated crop yield loss risks across different climate scenarios.
• To analyze the pathways through which climate factors influence crop yield via changes in crop growth processes.

Study Configuration

• Spatial Scale: Bengbu City, located in the middle reaches of the Huai River Basin, Anhui Province, China. Specifically, field experiments at Xinmaqiao Farmland Irrigation and Drainage Experiment Station (33°09′N, 117°22′E) and meteorological data from Bengbu National Meteorological Station (32°57′N, 117°22′E).
• Temporal Scale: Historical period: 1960–2019. Future period: 2021–2100 under SSP1–2.6, SSP2–4.5, and SSP5–8.5 scenarios. Daily data aggregated to 5-day intervals (pentads) for AFD identification.

Methodology and Data

• Models used:
  • DSSAT-CERES-Maize model: For simulating crop growth and yield formation processes. Calibrated using the Generalized Likelihood Uncertainty Estimation (GLUE) method.
  • BCC-CSM2-MR climate model: A global coupled climate model from the NEX-GDDP-CMIP6 dataset, used for future climate projections.
  • Gumbel copula: Applied to construct the joint distribution of AFD duration and severity for risk assessment.
  • Partial Least Squares Structural Equation Modeling (PLS-SEM): Employed to analyze impact pathways linking climatic conditions to maize yield.
  • Bayesian Information Criterion (BIC): Used for model selection of marginal distributions for AFD characteristics.
• Data sources:
  • Field experimental data (2016–2017): Soil and crop management data (sowing date, phenological dates, maturity date, grain yield, soil profile parameters) from Xinmaqiao Station.
  • Historical meteorological observations (1960–2019): Daily precipitation (P), maximum temperature (Tmax), minimum temperature (Tmin), and solar radiation (Ra) from Bengbu National Meteorological Station.
  • Future climate scenario data (2021–2100): Daily P, Tmax, Tmin, and Ra outputs from BCC-CSM2-MR under SSP1–2.6, SSP2–4.5, and SSP5–8.5 scenarios, bias-corrected using a univariate method.

Main Results

• Historical AFD events exhibited stage-dependent variation, with a mean annual cumulative severity of 9.54 (dimensionless) during 1960–2019, and the highest occurrence frequency in the milking stage (S4).
• Under the SSP5–8.5 scenario, AFD duration during the tasseling stage (S3) is projected to decrease by 8.33%, while severity is expected to increase by 3.44% around 2080, indicating an enhanced tendency toward flash-drought conditions.
• Yield losses varied significantly among AFD types, with events affecting the reproductive stages (S3 and S4) posing the highest risk. Historical yield reductions ranged from 13.98% to 39.80%, with future maximum reductions exceeding 80% under SSP5–8.5.
• The maize growing-season length is projected to shorten under future scenarios (e.g., by 23.6 days under SSP5–8.5), with the most significant reduction occurring during the milking stage (S4), decreasing by 36.82% under SSP5–8.5.
• Average maize yield is projected to decrease across all future scenarios, with approximately 67.50% of years exhibiting yield reductions exceeding 50% under SSP5–8.5, indicating a decline in yield stability (coefficient of variation = 1.12).
• Pathway analysis identified maximum temperature (Tmax) as the primary climatic driver of yield reduction, with the duration of the reproductive stages (S3 and S4) acting as key mediating variables. Evapotranspiration (ET) during S3 and S4 is projected to increase (24.48–29.85% for S3), intensifying water deficits.

Contributions

• Developed a novel crop-growth-process-based framework for agricultural flash drought (AFD) risk assessment that integrates crop growth stages with dynamic crop water supply–demand relationships.
• Introduced a drought identification time-step (DIT) based on 5-day intervals derived from the Twenty-four Solar Terms, improving the temporal resolution of agricultural drought monitoring compared to fixed-timescale indices.
• Quantified the duration, severity, joint occurrence risk, and yield impacts of AFD events under historical and future climate scenarios, specifically identifying the critical reproductive stages (S3 and S4) as most vulnerable.
• Utilized Partial Least Squares Structural Equation Modeling (PLS-SEM) to analyze the complex pathways through which climate factors, particularly maximum temperature, influence crop yield by mediating crop growth processes.
• Provided scientific and technical support for targeted irrigation scheduling and agricultural drought risk management by identifying key risk windows for AFD under multiple climate change scenarios.

Funding

• National Natural Science Foundation of China (52409017, 42271084, 52379006)
• Anhui Provincial Natural Science Foundation (2308085US13, 2408055US007)
• Sichuan Provincial Natural Science Foundation (2025ZNSFSC1369)
• Open Research Fund Program of State Key Laboratory of Eco-hydraulics in Northwest Arid Region, Xi’an University of Technology (2024KFKT-4)
• Open Research Fund Program of State Key Laboratory of Water Disaster Prevention (2025491411)

Citation

@article{Yanqi2026Agricultural,
  author = {Yanqi, Wei and Yuliang, Zhou and Juliang, Jin and Li, Haichao and bo, Ming and Zongzhi, WANG and Cui, Yi and Jin, TAO and Zhou, Ping},
  title = {Agricultural flash drought risk assessment under climate change based on crop growth stages},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.104158},
  url = {https://doi.org/10.1016/j.ejrh.2026.104158}
}