Remotely sensed prediction of soil organic carbon

Abstract

Soil organic carbon (SOC) is a critical indicator of soil health and plays a significant role in global carbon cycling. Based on the multiple roles that soil carbon plays in pasture ecosystems, such as reducing greenhouse gases, improving soil yield and fertility, reducing soil erodibility, and increasing water and food storage capacity, it is evident that carbon management is crucial to improving pasture quality. Remote sensing technologies have the potential to provide accurate and efficient predictions of SOC at large spatial scales, making it a valuable tool for monitoring and managing soil carbon stocks. Remote sensing can help inform policies and practices aimed at promoting sustainable land use by allowing for monitoring and management of soil carbon stocks at regional scales, The use of remote sensing for SOC prediction involves the integration of spectral, spatial, and temporal data. Spectral data, derived from remote sensing imagery, can provide information on vegetation cover, soil moisture, and other factors that influence SOC. Spatial data, such as topography and land use, can also be incorporated into SOC prediction models. Temporal data, including changes in vegetation cover and weather patterns, can provide insights into the dynamics of SOC. Algorithms, such as multivariate regression and factor analysis, have been successfully applied to predict SOC from remote sensing data. However, the accuracy of SOC predictions from remote sensing depends on factors such as the quality and frequency of the data, the selection of predictive variables, and the calibration/validation of the models. To improve the accuracy of SOC predictions, remote sensing can be integrated with other data sources, such as soil sampling. This integration can help to validate and calibrate the remote sensing models, ultimately leading to more accurate predictions of SOC. © 2024 Elsevier Inc. All rights reserved.