<p>From the perspective of the Sustainable Development Goals (SDGs), assessing the interaction between future food production potential and climate change is of paramount importance. Few studies have evaluated agricultural production systems across multiple scales and climate scenarios—an essential step toward understanding and optimizing future crop distribution. Therefore, this paper proposes a comprehensive framework that employs a geographically guided matrix to constrain the spatial distribution of cultivated crops and enhance fitting accuracy, enabling the model to focus on other key factors. Additionally, the Particle Swarm Optimization (PSO)-enhanced XGBoost method is utilized to explore the climatic and soil drivers of major crops (wheat, maize, and rice) and to establish predictive models for assessing crop response and adaptation to different climate change scenarios. The key findings are as follows: (1) The geographically guided matrix can effectively constrain the spatial distribution of production potential. Additionally, the PSO algorithm improves the test set accuracy by 0.7–31.7% and significantly mitigates overfitting at the municipal scale. (2) Wheat is the most sensitive to evapotranspiration stress. For maize and rice, their most sensitive stress factors exhibit scale-dependent differences but remain consistent across scales. (3) At a fine scale, there is a peak in the regulation of crop production potential by soil moisture, with values ranging around 0.25 <InlineEquation ID="IEq1"> <EquationSource Format="TEX">$({10}^{-2}y{r}^{-1})$</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mo stretchy="false">(</mo> <msup> <mrow> <mn>10</mn> </mrow> <mrow> <mo>-</mo> <mn>2</mn> </mrow> </msup> <mi>y</mi> <msup> <mrow> <mi>r</mi> </mrow> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mo stretchy="false">)</mo> </mrow> </math></EquationSource> </InlineEquation>. (4) Under future climate change scenarios, the high-yield regions for wheat are projected to shift northward to some extent.</p>

Visualisation of Key Thresholds of Crop Production Potential and Their Future Distribution Patterns in China Under Climate Change Scenarios

期刊论文
发表日期：

2025年8月1日

查看全文

Kui Yang,
Yanni Ma,
Qi Tang

摘要

From the perspective of the Sustainable Development Goals (SDGs), assessing the interaction between future food production potential and climate change is of paramount importance. Few studies have evaluated agricultural production systems across multiple scales and climate scenarios—an essential step toward understanding and optimizing future crop distribution. Therefore, this paper proposes a comprehensive framework that employs a geographically guided matrix to constrain the spatial distribution of cultivated crops and enhance fitting accuracy, enabling the model to focus on other key factors. Additionally, the Particle Swarm Optimization (PSO)-enhanced XGBoost method is utilized to explore the climatic and soil drivers of major crops (wheat, maize, and rice) and to establish predictive models for assessing crop response and adaptation to different climate change scenarios. The key findings are as follows: (1) The geographically guided matrix can effectively constrain the spatial distribution of production potential. Additionally, the PSO algorithm improves the test set accuracy by 0.7–31.7% and significantly mitigates overfitting at the municipal scale. (2) Wheat is the most sensitive to evapotranspiration stress. For maize and rice, their most sensitive stress factors exhibit scale-dependent differences but remain consistent across scales. (3) At a fine scale, there is a peak in the regulation of crop production potential by soil moisture, with values ranging around 0.25 $({10}^{-2}y{r}^{-1})$ $(10^{- 2} y r^{- 1})$ . (4) Under future climate change scenarios, the high-yield regions for wheat are projected to shift northward to some extent.