Scholarlink

<p>Dissolved organic carbon (DOC) in groundwater plays a pivotal role in regulating biogeochemical cycles and contaminant mobility within wetland critical zones. Despite its environmental significance, the spatiotemporal dynamics and controlling mechanisms of DOC remain poorly understood due to complex interactions among hydrological, hydrochemical, and environmental factors. Through monthly sampling campaigns (September 2022–February 2024) at the Xiangjiang River-Dongting Lake Wetland confluence, this study investigated groundwater DOC variability and driving factors using integrated machine learning models with SHapley Additive exPlanations (SHAP) and structural equation modeling (SEM). The results showed that significantly higher DOC concentrations in groundwater (2.81–48.50&#xa0;mg L⁻¹) compared to surface water (3.05–23.60&#xa0;mg L⁻¹). Groundwater DOC showed a seasonal pattern of decrease during the wet season and increase during the dry season. Spatial analysis revealed DOC concentrations decrease with increasing distance from the riverbank. The light gradient boosting machine (LightGBM) and categorical boosting (CatBoost) models outperformed other algorithms, demonstrating robust predictive capability for DOC variability. SHAP analysis identified NH₄⁺, COD<sub>Mn</sub>, temperature, and electrical conductivity as dominant drivers, while SEM revealed hydrochemical factors exerted the strongest direct control as they can directly reflect the biogeochemical processes between groundwater and aquifer. Meteorological and hydrological factors had a secondary influence through water table fluctuations, while soil factors showed minimal effect because their impact on DOC is typically mediated through indirect pathways involving hydrochemical factors. Our findings provide a quantitative framework for predicting DOC dynamics in groundwater and offer valuable insights for the protection and management of wetland ecosystems.</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

Spatiotemporal variability and driving factors of dissolved organic carbon in groundwater within the critical zone of wetland

  • 期刊论文

  • 发表日期：

    2025年7月1日

查看全文
  • Shuaishuai Lu,
  • Nianqing Zhou,
  • Yi Cai,
  • Mengshen Guo,
  • Kehao Liu,
  • Wengang Zhao

摘要

Dissolved organic carbon (DOC) in groundwater plays a pivotal role in regulating biogeochemical cycles and contaminant mobility within wetland critical zones. Despite its environmental significance, the spatiotemporal dynamics and controlling mechanisms of DOC remain poorly understood due to complex interactions among hydrological, hydrochemical, and environmental factors. Through monthly sampling campaigns (September 2022–February 2024) at the Xiangjiang River-Dongting Lake Wetland confluence, this study investigated groundwater DOC variability and driving factors using integrated machine learning models with SHapley Additive exPlanations (SHAP) and structural equation modeling (SEM). The results showed that significantly higher DOC concentrations in groundwater (2.81–48.50 mg L⁻¹) compared to surface water (3.05–23.60 mg L⁻¹). Groundwater DOC showed a seasonal pattern of decrease during the wet season and increase during the dry season. Spatial analysis revealed DOC concentrations decrease with increasing distance from the riverbank. The light gradient boosting machine (LightGBM) and categorical boosting (CatBoost) models outperformed other algorithms, demonstrating robust predictive capability for DOC variability. SHAP analysis identified NH₄⁺, CODMn, temperature, and electrical conductivity as dominant drivers, while SEM revealed hydrochemical factors exerted the strongest direct control as they can directly reflect the biogeochemical processes between groundwater and aquifer. Meteorological and hydrological factors had a secondary influence through water table fluctuations, while soil factors showed minimal effect because their impact on DOC is typically mediated through indirect pathways involving hydrochemical factors. Our findings provide a quantitative framework for predicting DOC dynamics in groundwater and offer valuable insights for the protection and management of wetland ecosystems.