<p>The complexity of farmland soil systems poses significant challenges in disentangling the interaction of environmental factors affecting heavy metal accumulation. Previous studies have predominantly focused on qualitatively characterizing the interactions between environmental factors on soil heavy metal concentrations in farmland, often neglecting in-depth quantitative analysis. In this study, a novel interaction effects quantization model based on collaborative filtering combining global spatial graph and local property graph is designed to analyze the interaction between the environmental factors and their potential impacts on soil Cd/As concentrations in farmland. The results revealed that the interaction effects of precipitation and river density influencing Cd/As accumulation exhibit nonlinear synergistic relationships, with R<sup>2</sup> values of 0.64 and 0.73, respectively. Moreover, the contributions of environmental factors on the interaction effects demonstrated dynamic trends: precipitation had the greatest contribution at moderate levels (958.9–964.4&#xa0;mm), while river density became more influential at higher levels (1.3–2.2). The interaction effects related to soil Cd showed no spatial disparity, whereas those associated with soil As depended strongly on the geological location. These findings offer valuable insights into the refined analysis of environmental factors influencing soil Cd/As concentrations in farmland and their interactions.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Quantify the environmental factors interaction affecting soil Cd/As accumulation in farmland using collaborative filtering model combining global spatial and local property graph

  • Renjie Zhang,
  • Yunhe Xie,
  • Tao Xue,
  • Saihua Liu,
  • Faxiang Tian,
  • Xionghui Ji,
  • Shufang Pan

摘要

The complexity of farmland soil systems poses significant challenges in disentangling the interaction of environmental factors affecting heavy metal accumulation. Previous studies have predominantly focused on qualitatively characterizing the interactions between environmental factors on soil heavy metal concentrations in farmland, often neglecting in-depth quantitative analysis. In this study, a novel interaction effects quantization model based on collaborative filtering combining global spatial graph and local property graph is designed to analyze the interaction between the environmental factors and their potential impacts on soil Cd/As concentrations in farmland. The results revealed that the interaction effects of precipitation and river density influencing Cd/As accumulation exhibit nonlinear synergistic relationships, with R2 values of 0.64 and 0.73, respectively. Moreover, the contributions of environmental factors on the interaction effects demonstrated dynamic trends: precipitation had the greatest contribution at moderate levels (958.9–964.4 mm), while river density became more influential at higher levels (1.3–2.2). The interaction effects related to soil Cd showed no spatial disparity, whereas those associated with soil As depended strongly on the geological location. These findings offer valuable insights into the refined analysis of environmental factors influencing soil Cd/As concentrations in farmland and their interactions.