Baluch et al. (2026) Evaluating Crop Yield Sensitivity to Climate Change Using a Deep‐Learning Framework Incorporating Vegetation and Groundwater Memory

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Earth s Future
• Year: 2026
• Date: 2026-08-28
• Authors: Shehakk Muneer Baluch, Muhammad Abrar Faiz, Luchen Wang, Wuyuan Liu, Haiyan Li, Mo Li
• DOI: 10.1029/2026ef008560

Research Groups

• Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences (CAS)
• Key Laboratory of Land Surface Process and Climate Change, CAS
• School of Earth Sciences, Peking University

Short Summary

This study developed an interpretable spatiotemporal framework to reconstruct historical crop yields responses in Northeast China and assess future climate-scenario impacts. The framework used a combination of machine learning models and satellite data to predict crop yields with high accuracy.

Objective

• Investigate the impact of eco-hydrological memory on crop yields in Northeast China

Study Configuration

• Spatial Scale: Northeast China (NEC), a region critical for national food security
• Temporal Scale: 1981-present, with future projections from 2017 to 2060

Methodology and Data

• Models used:
  • ReGen-Net: a reconstruction module that extends vegetation Normalized Difference Vegetation Index and groundwater storage-anomaly memory back to 1981
  • GATAE-TFT: a spatially informed model for scenario-based projections, achieving test R2 values of 0.857, 0.777, 0.798, and 0.837 for maize, rice, soybean, and wheat, respectively
  • YieldTransformer: a temporal baseline model
• Data sources: Satellite data (e.g., NDVI), groundwater storage-anomaly memory, and four decades of yield records

Main Results

• The study reconstructed historical crop yields responses in NEC with high accuracy, using the GATAE-TFT model.
• Future climate-scenario projections indicate gradual declines in maize, rice, and wheat yields, while soybean yields remain stable under both SSP1-2.6 and SSP5-8.5.

Contributions

• This study provides a physically grounded and interpretable benchmark for assessing water-stress risks to crop production in NEC.
• The framework developed in this study can be applied to other regions with similar eco-hydrological characteristics.

Funding

• National Natural Science Foundation of China (Grant No. 41921003)
• Key Research Program of Frontier Sciences, CAS (Grant No. QYZDY-SSW-DQC001)

Citation

@article{Baluch2026Evaluating,
  author = {Baluch, Shehakk Muneer and Faiz, Muhammad Abrar and Wang, Luchen and Liu, Wuyuan and Li, Haiyan and Li, Mo},
  title = {Evaluating Crop Yield Sensitivity to Climate Change Using a Deep‐Learning Framework Incorporating Vegetation and Groundwater Memory},
  journal = {Earth s Future},
  year = {2026},
  doi = {10.1029/2026ef008560},
  url = {https://doi.org/10.1029/2026ef008560}
}