Scholarlink

Purpose <p>Total organic matter (TOM) in lake and marsh sediments is an important indicator of ecosystem function, carbon storage, and environmental changes. This study evaluated whether reflectance spectroscopy and color coefficients can provide accurate, rapid, and cost-effective alternatives to conventional TOM quantification in sediment cores.</p> Materials and methods <p>A total of 350 samples from 10 sediment cores were collected in Netley Marsh, a freshwater coastal wetland at the southern end of Lake Winnipeg, Manitoba, Canada. Reflectance spectra were obtained across the visible (VIS; 350–700&#xa0;nm), near-infrared (NIR; 700–1000&#xa0;nm), short-wave infrared (SWIR; 1000–2500&#xa0;nm), and full-spectrum (350–2500&#xa0;nm) ranges. Fifteen color coefficients were derived from the visible range. Partitioning Around Medoids (PAM) clustering was used to explore spectral shape variation, while Partial Least Squares Regression (PLSR), Random Forest (RF), and Cubist models were trained using 10-fold cross-validation and validated on an independent test set.</p> Results <p>All models achieved strong predictive performance (R² &gt; 0.80, RPD &gt; 2.2). Cubist consistently outperformed the other models, with the visible range yielding the most accurate results (R² = 0.95, RMSE = 1.7%, RPD = 4.32). Models based on color coefficients performed comparably to full-spectrum inputs.</p> Conclusion <p>Visible-range reflectance and derived color metrics provide robust, rapid, and cost-efficient prediction of TOM in sediment cores. This approach offers a practical tool for reconstructing carbon dynamics and supporting sediment and wetland management in aquatic 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

Prediction of total organic matter in marsh sediments: integrating reflectance clustering, spectral subranges, and color coefficients

  • 期刊论文

  • 发表日期：

    2026年5月30日

查看全文
  • Z. A. Dastgheib,
  • M. Goharrokhi,
  • S. Clark,
  • D. Lobb,
  • N. Badreldin,
  • G. McCullough

摘要

Purpose

Total organic matter (TOM) in lake and marsh sediments is an important indicator of ecosystem function, carbon storage, and environmental changes. This study evaluated whether reflectance spectroscopy and color coefficients can provide accurate, rapid, and cost-effective alternatives to conventional TOM quantification in sediment cores.

Materials and methods

A total of 350 samples from 10 sediment cores were collected in Netley Marsh, a freshwater coastal wetland at the southern end of Lake Winnipeg, Manitoba, Canada. Reflectance spectra were obtained across the visible (VIS; 350–700 nm), near-infrared (NIR; 700–1000 nm), short-wave infrared (SWIR; 1000–2500 nm), and full-spectrum (350–2500 nm) ranges. Fifteen color coefficients were derived from the visible range. Partitioning Around Medoids (PAM) clustering was used to explore spectral shape variation, while Partial Least Squares Regression (PLSR), Random Forest (RF), and Cubist models were trained using 10-fold cross-validation and validated on an independent test set.

Results

All models achieved strong predictive performance (R² > 0.80, RPD > 2.2). Cubist consistently outperformed the other models, with the visible range yielding the most accurate results (R² = 0.95, RMSE = 1.7%, RPD = 4.32). Models based on color coefficients performed comparably to full-spectrum inputs.

Conclusion

Visible-range reflectance and derived color metrics provide robust, rapid, and cost-efficient prediction of TOM in sediment cores. This approach offers a practical tool for reconstructing carbon dynamics and supporting sediment and wetland management in aquatic ecosystems.