An application of vis-NIR reflectance spectroscopy and Artificial Neural Networks to the prediction of soil organic carbon content in southern Italy

Understanding soil properties is an essential prerequisite for sustainable land management. Assessment of these properties has long been gained through conventional laboratory analysis, which is considered costly and time consuming. Therefore, there is a need to develop alternative cheaper and faster techniques for soil analysis. In recent years, special attention has been given to vis-NIR reflectance spectroscopy and chemometrics. In this study we evaluated the potential of vis-NIR spectroscopy and Back Propagation Neural Networks (BPNN) for prediction of organic carbon (OC) of soils representative of three Mediterranean agro-ecosystems from the Campania region, southern Italy. An Artificial Neural Network (ANN) model was developed based on Multi-Layer Perceptron (MLP) network and trained by a Back-Propagation algorithm on reflectance data. The training and validation phases, confirmed by a ten fold cross validation methodology, led to a very satisfactory calibration of the BPNN model.