Scholarlink

<p>Reliable geochemical baselines are largely absent for northern Ghana, limiting efforts to distinguish natural element variability from human-induced contamination. This study addresses that gap by evaluating soil geochemical compositions in the Bongo and Talensi districts, where limited prior characterization has hindered accurate environmental assessment. Using an integrated geostatistical and machine-learning framework, regional background and baseline values were established to support environmental monitoring, land-use planning, and resource management. Three complementary geostatistical approaches; Iterative (3<i>σ</i>), Frequency Distribution, and Concentration–Area (C–A), were combined with clustering and regression-based learning to delineate geochemical zones and identify key elemental drivers. Machine-learning analysis identified Cd, Sb, Ge, and Ag as the main predictors distinguishing the Bongo and Talensi geochemical provinces. The Bongo province, underlain by felsic granitoids, is dominated by silicate weathering and low trace-metal variability, whereas Talensi reflects metavolcanic and hydrothermally influenced soils with localized metal enrichment. Chromium (Cr) exhibited the highest mean concentration (276.6&#xa0;mg/kg), exceeding international soil-quality limits, while Cu (7.9&#xa0;mg/kg), Pb (4.8&#xa0;mg/kg), and Zn (26.2&#xa0;mg/kg) remained within safe thresholds. The enrichment of Cr and related baseline ratios indicate that trace-metal variations arise chiefly from natural bedrock composition. The results provide reliable geochemical baselines essential for contamination assessment, environmental regulation, and mineral exploration. This study delivers the first integrated geostatistical–machine-learning framework for northern Ghana, offering a practical tool for sustainable land-use and resource-management decision-making.</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

Soil geochemistry and contamination zoning in Northeastern Ghana: insights from the Bongo and Talensi districts

  • 期刊论文

  • 发表日期：

    2026年1月8日

查看全文
  • Belinda S. Berdie,
  • Raymond W. Kazapoe,
  • Darwin A. Awog-Badek,
  • Blestmond A. Brako,
  • Gordon Foli,
  • Simon K. Y. Gawu

摘要

Reliable geochemical baselines are largely absent for northern Ghana, limiting efforts to distinguish natural element variability from human-induced contamination. This study addresses that gap by evaluating soil geochemical compositions in the Bongo and Talensi districts, where limited prior characterization has hindered accurate environmental assessment. Using an integrated geostatistical and machine-learning framework, regional background and baseline values were established to support environmental monitoring, land-use planning, and resource management. Three complementary geostatistical approaches; Iterative (3σ), Frequency Distribution, and Concentration–Area (C–A), were combined with clustering and regression-based learning to delineate geochemical zones and identify key elemental drivers. Machine-learning analysis identified Cd, Sb, Ge, and Ag as the main predictors distinguishing the Bongo and Talensi geochemical provinces. The Bongo province, underlain by felsic granitoids, is dominated by silicate weathering and low trace-metal variability, whereas Talensi reflects metavolcanic and hydrothermally influenced soils with localized metal enrichment. Chromium (Cr) exhibited the highest mean concentration (276.6 mg/kg), exceeding international soil-quality limits, while Cu (7.9 mg/kg), Pb (4.8 mg/kg), and Zn (26.2 mg/kg) remained within safe thresholds. The enrichment of Cr and related baseline ratios indicate that trace-metal variations arise chiefly from natural bedrock composition. The results provide reliable geochemical baselines essential for contamination assessment, environmental regulation, and mineral exploration. This study delivers the first integrated geostatistical–machine-learning framework for northern Ghana, offering a practical tool for sustainable land-use and resource-management decision-making.