Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado

Author(s)

Canteral, Kleve Freddy Ferreira ; Vicentini, Maria Elisa ; de Lucena, Wanderson Benerval ; de Moraes, Mario Luiz Teixeira ; Montanari, Rafael ; Ferraudo, Antonio Sergio ; Peruzzi, Nelson Jose ; La Scala Jr, Newton ; Panosso, Alan Rodrigo

• Document type: Journal article
• Source: Environmental Science and Pollution Research; v. 30, n. 21, p. 20-pg., 2023-04-12.

Abstract

Soil CO2 emission (FCO2) is a critical component of the global carbon cycle, but it is a source of great uncertainty due to the great spatial and temporal variability. Modeling of soil respiration can strongly contribute to reducing the uncertainties associated with the sources and sinks of carbon in the soil. In this study, we compared five machine learning (ML) models to predict the spatiotemporal variability of FCO2 in three reforested areas: eucalyptus (RE), pine (RP) and native species (RNS). The study also included a generalized scenario (GS) where all the data from RE, RP and RNS were included in one dataset. The ML models include generalized regression neural network (GRNN), radial basis function neural network (RBFNN), multilayer perceptron neural network (MLPNN), adaptive neuro-fuzzy inference system (ANFIS) and random forest (RF). Initially, we had 32 attributes and after pre-processing, including Pearson's correlation, canonical correlation analysis (CCA), and biophysical justification, only 21 variables remained. We used as input variables 19 soil properties and climate variables in reforested areas of eucalyptus, pine and native species. RF was the best model to predict soil respiration to RE [adjusted coefficient of determination (R-2 adj): 0.70 and root mean square error (RMSE): 1.02 mu mol m(-2) s(-1)], RP (R-2 adj: 0.48 and RMSE: 1.07 mu mol-m(-2) s(-1)) and GS (R-2 adj: 0.70 and RMSE: 1.05 mu mol m(-2) s(-1)). Our findings support that RF and GRNN are promising for predicting soil respiration of reforested areas which could help to identify and monitor potential sources and sinks of the main additional greenhouse gas over ecosystems. (AU)

• FAPESP's process: 16/03861-5 - Soil CO2 emission and carbon stocks in agricultural areas and planted forests in the cerrado region of Mato Grosso do Sul
• Grantee: Alan Rodrigo Panosso
• Support Opportunities: Research Program on Global Climate Change - Regular Grants
• FAPESP's process: 08/58187-0 - Impact of management practices on soil CO2 emission in sugarcane production areas, Southern Brazil
• Grantee: Newton La Scala Júnior
• Support Opportunities: Research Program on Global Climate Change - Thematic Grants