Prediction of Soil Organic Carbon and its Physical Fractions Using Near-Infrared Spectroscopy and Chemometric Modeling in Tropical Savannas of the Colombian Eastern Plains

- 335857
Posters
Favorite this paper
How to cite this paper?
Abstract

Soil organic carbon (SOC) is a key indicator of soil quality and ecosystem processes in the Colombian Eastern Plains. Conventional SOC quantification is labor-intensive and costly, which constrains large-scale monitoring efforts. This study evaluated the potential of combining near-infrared (NIR) spectroscopy and chemometric techniques to predict SOC at AGROSAVIA's La Libertad Research Center (Meta, Colombia). Using HCLC, twenty-three soil profiles (100 cm depth) were selected and subdivided into six depth intervals (0–10, 10–20, 20–30, 30–50, 50–70, and 70–100 cm), followed by physical fractionation through sieving at 2 mm, 75, 63, and 25 μm. Organic carbon was determined by the Walkley–Black method, and NIR spectra were acquired for each subsample. Four calibration approaches (SNV, LASSO, PLSR, and Random Forest) were developed and compared using R², RMSE, and RPD as performance metrics. Model performance varied according to the algorithm employed, reflecting the inherent complexity of carbon distribution among soil fractions. The SVR model exhibited the best predictive model for SOC estimation (R² = 0.71, RMSE = 0.35 g kg⁻¹, RPD = 1.85). Depth-stratified calibration improved model accuracy relative to global models. Overall, this study demonstrates that NIR spectroscopy remains a rapid and cost-effective alternative for assessing SOC and its fractions in tropical savanna ecosystems, contributing to sustainable land management and carbon monitoring in the Eastern Plains region.

Share your ideas or questions with the authors!

Did you know that the greatest stimulus in scientific and cultural development is curiosity? Leave your questions or suggestions to the author!

Sign in to interact

Have a question or suggestion? Share your feedback with the authors!

Institutions
  • 1 Universidad Nacional de Colombia
  • 2 AGROSAVIA
Track
  • SOM modeling in agricultural and natural ecosystems
Keywords
Soil carbon fractions
Tropical savannas
Depth stratification
Predictive modeling
Support vector regression (SVR)