Scholarlink

<p>The changes in soil temperature and moisture due to freezing and thawing can modify the carbon and nitrogen release rates from soils in ecological buffer zones, affecting the carbon-nitrogen balance of lake wetlands in arid regions. This study selected the ecological buffer zone of saltwater lakes in arid areas as the research subject. Through in - situ monitoring and simulation via the Generalized Regression Neural Network (GRNN) model, the regulatory mechanisms of soil hydrothermal dynamics on CO₂, CH₄, and N₂O emissions were revealed. The results showed that: (1) The buffer zone overall acted as a CO₂ emission source (98,984.90&#xa0;mg·m⁻²·h⁻¹) and a CH₄ absorption sink (− 6.36&#xa0;mg·m⁻²·h⁻¹), where soils in aquatic habitats contributed 80.48% of CO₂ emissions, and CH₄ oxidation predominantly occurred in shallow soils (0–20&#xa0;cm). (2) N₂O emissions (3.02 mg·m<sup>− 2</sup>·h<sup>− 1</sup>) decreased along the habitat gradient and shifted to uptake in sandy habitats. (3) The GRNN model identified key thresholds: when soil moisture content ranges from 1 to 8.4%, CO₂ emission peaks. CH₄ emission peaks at soil temperatures of 26–27.2℃ and soil moisture content of 4.2–16.6%. N₂O emission reaches its maximum at 15.5–24.4℃ and with moisture content below 9%. These findings highlight the role of soil hydrothermal changes in wetland carbon-nitrogen cycling, as well as the importance of soil moisture in regulating carbon emissions under a warmer, drier climate.</p> Graphical Abstract <p>This graphical abstract presents research on the Ebinur Lake wetland buffer zone, a key area in arid ecosystems, integrating in-situ monitoring data from sandy, saline, and aquatic habitats. It interprets soil carbon and nitrogen emissions through analysis combining field observations with Generalized Regression Neural Network (GRNN) modeling. Key components include: data sources from in-situ measurements of CO₂, CH₄, N₂O fluxes and soil hydrothermal dynamics across habitat gradients, covering 144 sets of annual data from March to February; model application involving the integration of these 144 datasets—via the GRNN model—to explore the response relationship between hydrothermal conditions (with 1% changes) and soil greenhouse gases (CO₂, CH₄, N₂O), alongside the identification of critical hydrothermal thresholds; core results revealing significant correlations between greenhouse gas emissions from soils in different vegetation habitats and hydrothermal conditions: aquatic habitats dominate CO₂ emissions (80.48%), shallow soils (0–20&#xa0;cm) drive CH₄ absorption, N₂O fluxes decrease along the aquatic→saline→sandy habitat gradient (with reverse absorption in sandy habitats), and the GRNN model identifies peak CO₂ emissions at 1–8.4% soil moisture, peak CH₄ emissions at 26–27.2℃ with 4.2–16.6% moisture, and peak N₂O emissions at 15.5–24.4℃ with &lt; 9% moisture; conclusions indicating that hydrothermal dynamics regulate carbon-nitrogen cycles, with higher soil moisture critical for emission reduction under climate warming. This synthesis demonstrates how multi-habitat monitoring and machine learning clarify wetland biogeochemical responses, providing a basis for arid region climate mitigation strategies.</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

Freeze-Thaw-Driven Multi-scale Quantification of Hydrothermal Thresholds for GHG Emissions in Arid-Region Saltwater Lake Buffer Zones Via GRNN

  • 期刊论文

  • 发表日期：

    2026年3月5日

查看全文
  • Guanghua Miao,
  • Yanhong Li,
  • Yu Zhu,
  • Shujing Yin,
  • Aerdake Aqing

摘要

The changes in soil temperature and moisture due to freezing and thawing can modify the carbon and nitrogen release rates from soils in ecological buffer zones, affecting the carbon-nitrogen balance of lake wetlands in arid regions. This study selected the ecological buffer zone of saltwater lakes in arid areas as the research subject. Through in - situ monitoring and simulation via the Generalized Regression Neural Network (GRNN) model, the regulatory mechanisms of soil hydrothermal dynamics on CO₂, CH₄, and N₂O emissions were revealed. The results showed that: (1) The buffer zone overall acted as a CO₂ emission source (98,984.90 mg·m⁻²·h⁻¹) and a CH₄ absorption sink (− 6.36 mg·m⁻²·h⁻¹), where soils in aquatic habitats contributed 80.48% of CO₂ emissions, and CH₄ oxidation predominantly occurred in shallow soils (0–20 cm). (2) N₂O emissions (3.02 mg·m− 2·h− 1) decreased along the habitat gradient and shifted to uptake in sandy habitats. (3) The GRNN model identified key thresholds: when soil moisture content ranges from 1 to 8.4%, CO₂ emission peaks. CH₄ emission peaks at soil temperatures of 26–27.2℃ and soil moisture content of 4.2–16.6%. N₂O emission reaches its maximum at 15.5–24.4℃ and with moisture content below 9%. These findings highlight the role of soil hydrothermal changes in wetland carbon-nitrogen cycling, as well as the importance of soil moisture in regulating carbon emissions under a warmer, drier climate.

Graphical Abstract

This graphical abstract presents research on the Ebinur Lake wetland buffer zone, a key area in arid ecosystems, integrating in-situ monitoring data from sandy, saline, and aquatic habitats. It interprets soil carbon and nitrogen emissions through analysis combining field observations with Generalized Regression Neural Network (GRNN) modeling. Key components include: data sources from in-situ measurements of CO₂, CH₄, N₂O fluxes and soil hydrothermal dynamics across habitat gradients, covering 144 sets of annual data from March to February; model application involving the integration of these 144 datasets—via the GRNN model—to explore the response relationship between hydrothermal conditions (with 1% changes) and soil greenhouse gases (CO₂, CH₄, N₂O), alongside the identification of critical hydrothermal thresholds; core results revealing significant correlations between greenhouse gas emissions from soils in different vegetation habitats and hydrothermal conditions: aquatic habitats dominate CO₂ emissions (80.48%), shallow soils (0–20 cm) drive CH₄ absorption, N₂O fluxes decrease along the aquatic→saline→sandy habitat gradient (with reverse absorption in sandy habitats), and the GRNN model identifies peak CO₂ emissions at 1–8.4% soil moisture, peak CH₄ emissions at 26–27.2℃ with 4.2–16.6% moisture, and peak N₂O emissions at 15.5–24.4℃ with < 9% moisture; conclusions indicating that hydrothermal dynamics regulate carbon-nitrogen cycles, with higher soil moisture critical for emission reduction under climate warming. This synthesis demonstrates how multi-habitat monitoring and machine learning clarify wetland biogeochemical responses, providing a basis for arid region climate mitigation strategies.