Han et al. (2025) Spatiotemporal changes in agricultural planting structure in the Turpan–Hami Basin, Xinjiang, China: Remote sensing monitoring from 1990 to 2023

Identification

• Journal: Smart Agricultural Technology
• Year: 2025
• Date: 2025-11-16
• Authors: Xuemei Han, Huichun Ye, Xue Jiao, Chaojia Nie, Aynur Mamat, Mingyao Tang
• DOI: 10.1016/j.atech.2025.101626

Research Groups

• Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing, China
• International Research Center of Big Data for Sustainable Development Goals, Beijing, China
• College of Geology and Mining Engineering, Xinjiang University, Urumqi, China
• Lab of Big Earth Data and Sustainable Development Goal (BASL), Kashi Aerospace Information Research Institute, Kashgar, China
• Xinjiang Soil and Fertilizer Station, Urumqi, China

Short Summary

This study developed a remote sensing method using long-term Landsat imagery and crop phenology to monitor spatiotemporal changes in agricultural planting structure in the Turpan-Hami Basin from 1990 to 2023, revealing a significant expansion of total cultivated area and a distinct shift towards high-value economic crops.

Objective

• To characterize the spatiotemporal dynamics of agricultural planting structure in the Turpan-Hami Basin from 1990 to 2023 using remote sensing.
• To develop a novel time-series Extrema-Visual Separability (EVS) method for extracting sensitive remote sensing features, enhancing feature selection interpretability and classification accuracy.
• To construct a Random Forest-based crop identification model and generate a high-precision agricultural planting structure dataset for the region.
• To provide scientific support for agricultural resource optimization and sustainable development in arid oasis agricultural systems.

Study Configuration

• Spatial Scale: Turpan–Hami Basin, Xinjiang, China (encompassing Turpan and Hami cities, specifically the Turpan Basin, Hami Basin, and Bayin Basin).
• Temporal Scale: 1990 to 2023 (specifically for the years 1990, 1995, 2000, 2005, 2010, 2015, 2020, and 2023), focusing on the crop growing season from March to October each year.

Methodology and Data

• Models used:
  • Temporal Extreme-Value and Visual Separability (EVS) method for sensitive remote sensing feature extraction.
  • Random Forest (RF) algorithm for crop classification.
  • Savitzky-Golay (SG) filter for time-series noise removal.
  • Gray-Level Co-occurrence Matrix (GLCM) for texture feature calculation.
• Data sources:
  • Satellite imagery: Long-term Landsat series imagery (Landsat-5 TM and Landsat-8 OLI) Level-2 Surface Reflectance (SR) data, monthly composited.
  • Field sample data: 1600 ground sample points collected in June-July 2023, visually inspected using high-resolution Google Earth imagery, and transferred to previous years (1990-2020).
  • Auxiliary data: Crop phenological information, historical statistical data from the Xinjiang Statistical Yearbook for validation.

Main Results

• The proposed EVS method and Random Forest model achieved high and stable classification performance, with overall accuracies exceeding 85 % and Kappa coefficients consistently above 0.79 across all years.
• Comparison with statistical yearbook data showed minimal relative errors (within ±0.18 %) and strong consistency (R² consistently above 0.78).
• From 1990 to 2023, the total agricultural planting area in the Turpan-Hami Basin expanded from 104.95 × 10³ hectares to 191.90 × 10³ hectares, an increase of approximately 83 %.
• The regional cropping structure underwent a distinct transformation, shifting from traditional grain cultivation towards high-value economic crops.
• As of 2023, economic crops (grapes, cotton, melons, apricot trees, jujube trees) accounted for 81.85 % of the total planted area, with grapes (29.49 %), cotton (26.85 %), and melons (14.23 %) being the dominant crops.
• Grapevine cultivation showed the most significant increase, expanding by 277 % from 15.00 × 10³ hectares to 56.60 × 10³ hectares. Jujube and apricot trees also demonstrated rapid expansion.
• Conversely, the planting areas of food crops such as sorghum and wheat showed an overall decline, reflecting a structural shift towards higher-value economic crops.

Contributions

• Generated a robust, long-term (1990-2023), high-precision dataset of agricultural planting structures for the Turpan-Hami Basin using Landsat imagery.
• Introduced the novel Extrema–Visual Separability (EVS) method for selecting sensitive remote sensing features, enhancing interpretability and classification accuracy in multi-temporal crop monitoring.
• Provided a comprehensive spatiotemporal analysis of cropping structure evolution in an arid oasis region, revealing a significant shift towards high-value economic crops driven by policy, market demand, and improved irrigation.
• Offers a robust data foundation and methodological reference for scientific agricultural management and sustainable development in arid and semi-arid regions.

Funding

• Third Xinjiang Scientific Expedition Program of China (Grant No 2022xjkk1100)
• Natural Science Foundation of Xinjiang Uygur Autonomous Region, China (2024D01A21)
• Xinjiang Uygur Autonomous Region “Tianshan Talents” Talent Program-Three Rural Backbone Talent Project (No. 2024SNGGGCC043)

Citation

@article{Han2025Spatiotemporal,
  author = {Han, Xuemei and Ye, Huichun and Jiao, Xue and Nie, Chaojia and Mamat, Aynur and Tang, Mingyao},
  title = {Spatiotemporal changes in agricultural planting structure in the Turpan–Hami Basin, Xinjiang, China: Remote sensing monitoring from 1990 to 2023},
  journal = {Smart Agricultural Technology},
  year = {2025},
  doi = {10.1016/j.atech.2025.101626},
  url = {https://doi.org/10.1016/j.atech.2025.101626}
}