Crop Suitability

Abstract

Hydrothermal heterogeneity across China shapes macro-scale crop distribution, and modern agricultural technology increasingly blurs the traditional agricultural-pastoral boundary. We present a GIS-based qualitative and quantitative analysis to assess the climatic suitability of 7 major crops: rice, wheat, rapeseed, peanut, cotton, sugarcane, and sugar beet, and to analyze the spatial dynamics of the agricultural-pastoral ecotone.

Using polygon-based and point-based crop distribution data with 8 environmental rasters, we construct a three-metric adaptability framework (niche breadth, environmental elasticity, environmental typicality). Gaussian KDE generates a continuous crop richness surface; Random Forest with permutation importance identifies dominant drivers; Logistic regression with a transgression index quantifies each crop's penetration into pastoral zones.

Key findings: (1) DEM dominates crop richness, surpassing precipitation and temperature; (2) cotton, wheat, rice, and rapeseed rank highest in adaptability, while peanut ranks lowest due to strict soil-moisture requirements; (3) richness hotspots concentrate in the Sichuan Basin, the Yangtze and North China Plains, and Xinjiang's Yarkant Oasis, peaking at ~4.6; (4) sugar beet and cotton show the strongest transgression into pastoral zones, driven by irrigation and favorable diurnal temperature ranges. These findings provide quantitative reference for crop layout optimization and agricultural-pastoral policy under climate change.

Analysis Pipeline — Overall analysis pipeline.

Study Area & Data

We analyze 7 major crops (rice, wheat, rapeseed, peanut, cotton, sugarcane, sugar beet) across China, using concentrated (polygon) and scattered (point) distribution data paired with 8 environmental rasters spanning climate, terrain, and hydrology at unified resolution. The original data is on ModelScope.

DEM of China — Study area: Land of China.

Environmental Factors

8 gridded environmental factors used in the analysis.

Mean Temperature — Mean annual temperature (1901-2024).

Annual Precipitation — Mean annual precipitation (1901-2024).

Accumulated GDD — Mean annual accumulated growing degree days (1950-2024).

DEM — Digital elevation model (Copernicus 30 m).

Terrain Roughness — Terrain roughness derived from DEM.

Topographic Wetness Index — Topographic wetness index.

Distance to Rivers — Euclidean distance to nearest river.

Crop Richness

Continuous crop species richness surface generated via Gaussian KDE, with maximum richness reaching ~4.6.

Adaptability Ranking

Three-metric assessment (niche breadth, environmental elasticity, environmental typicality) ranking of 7 crops.

Suitability Prediction

Nationwide suitability prediction for each crop based on Logistic Regression models.

Agricultural-Pastoral Transgression

Relative transgression index quantifying each crop's penetration into pastoral zones.

License

This project is licensed under the Apache License 2.0. See the LICENSE file for details.

Citation

If you find our work or data useful in your research, please cite:

@misc{tang2026chinaagroeco,
  author    = {Tang, Haojun},
  title     = {{ChinaAgroEco}: A Multi-Source Agro-Ecological Dataset for China},
  year      = {2026},
  publisher = {ModelScope},
  howpublished = {\url{https://www.modelscope.cn/datasets/Donald123456/ChinaAgroEco}},
}

@misc{tang2026cropsuit,
  author    = {Tang, Haojun},
  title     = {Climate Suitability of Major Crop Distribution and Agricultural-Pastoral Zoning in China},
  year      = {2026},
  publisher = {GitHub},
  howpublished = {\url{https://github.com/DonaldTrump-coder/china-crop-climatic-suitability}},
}

Spatial Analysis:

Climate Suitability of Major Crop Distribution and Agricultural-Pastoral Zoning in China