Estimation of soil organic carbon content using remote sensing and GIS techniques

Abstract

The depletion of Soil Organic Carbon (SOC) due to intensive agricultural practices poses a significant threat to soil health, impacting agricultural productivity, soil structure, and carbon sequestration. Remote methods to evaluate the surface SOC content will enhance efficient mapping and therefore, apply appropriate methods for remediation. A research study was developed to provide a cost-effective, non-invasive method for SOC estimation and mapping, contributing to sustainable agriculture and environmental conservation. The focus of the study included using remote sensing (RS), satellite imagery and Geographic Information Systems (GIS) software to estimate soil organic carbon (SOC) content through various vegetation indices (VIs). Statistical analysis included both descriptive statistics and multivariate analyses. The SOC data did not follow a normal distribution, necessitating the use of non-parametric tests. The study employed multivariate correlation, Spearman's rho as non-parametric tests, and ordinal logistic regression to create SOC estimation models. The transformation of SOC data into ordinal classes allowed for more robust regression analysis, improving the predictive power of the models. The results showed significant correlations between SOC and the VIs, particularly with NDVI, GNDVI, and SAVI, with correlation coefficients above 0.9, indicating strong predictive capabilities. BSI exhibited an inverse relationship with SOC, as expected. The distribution analyses of the indices highlighted varying vegetation health and density across the study area, confirming the suitability of these indices for SOC estimation. The study underscored the potential of RS and GIS technologies in providing reliable SOC estimates, promoting Precision Agriculture (PA) and sustainable land management. It also suggests further refinement and validation of the models using Unmanned Aerial Systems (UASs) equipped with multispectral cameras of high resolution, to enhance spatial resolution and accuracy. Additionally, future research should explore the integration of more environmental variables and advanced statistical techniques to improve SOC prediction models. In conclusion, the utilization of RS and GIS for estimating SOC through VIs a promising avenue for enhancing soil management and conservation efforts. By leveraging advanced technologies and statistical methods, this study provides valuable insights into the complex interactions between vegetation and soil carbon dynamics, paving the way for more effective and sustainable agricultural practices. The study is under further validation in the same and other areas.