Modelling carbon dioxide emissions under a maize-soy rotation using machine learning

Machine learning (ML) models are an effective and efficient alternative to mechanistic models for predicting CO2 emissions from agricultural soils. Random forest (RF), a classical regression ML model, is a suitable algorithm to predict soil CO2 emissions regardless of fertiliser scenario. Feed-forward neural network (FNN) provides acceptable predictive performance for CO2 emissions, but it does not provide consistent predictive performance in K-Fold cross-validation. Climatic parameters influence CO2 emissions and the complexity of the relationship is not fully captured in biophysical models. Machine learning (ML) is now being applied to environmental problems, and it is, therefore, opportune to investigate ML models in CO2 predictions from agricultural soils.