Advancing soil carbon estimation: Innovative technologies for sustainable agriculture and climate action

Agriculture faces increasing pressure to adopt innovative technologies that address the dual challenge of enhancing productivity while mitigating the effects of global climate change. Proximal sensors have emerged as promising tools, enabling sustainable land management by optimizing resources, reducing environmental impacts, and aiding in precise soil characterization. Soil carbon (C) stocks accurate quantification is pivotal for promoting sustainable practices and supporting carbon sequestration efforts aligned with climate goals. This study aimed to assess the potential of diffuse reflectance spectroscopy (DRS) and magnetic susceptibility (MS) as predictive tools for estimating soil carbon stocks within the sandstone-basaltic transition zone, providing insights into their application for advancing global carbon sequestration strategies. This research used a soil carbon stock database (n = 446) from an area of ​​approximately 900 ha, located in the São Paulo state in Guatapara City, Brazil. Partial least squares regression (PLSR) was applied to the DRS data, while linear regression was used for MS data. The ParLeS software was used to perform PLSR analysis, establishing a relationship between the spectra and soil carbon. Prediction accuracy was evaluated based on correlation coefficients (r) and root mean square error (RMSE). Both methods demonstrated strong prediction accuracies, with DRS achieving r = 0.87 and RMSE = 1.24, and MS achieving r = 0.88 and RMSE = 1.98. However, carbon stock maps revealed overestimations of 5.6 t ha⁻¹ for DRS (28% error) and 9.2 t ha⁻¹ for MS (46% error). The results confirm the potential of DRS and MS as reliable techniques for estimating soil carbon stocks. Despite prediction errors, their alignment with certification protocols highlights their practicality for immediate application, with opportunities for refinement. This study underscores the importance of improving DRS and MS methodologies to enhance their accuracy and reliability in carbon stock estimation. Such advancements are critical for their successful integration into certified soil carbon sequestration programs, thereby contributing to climate change mitigation strategies and supporting Sustainable Development Goal 13 (Climate Action) of the 2030 Agenda.