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

Detalhes bibliográficos
Autor(a) principal: Canteral, Kleve Freddy Ferreira [UNESP]
Data de Publicação: 2023
Outros Autores: Vicentini, Maria Elisa [UNESP], de Lucena, Wanderson Benerval [UNESP], de Moraes, Mário Luiz Teixeira [UNESP], Montanari, Rafael [UNESP], Ferraudo, Antonio Sergio [UNESP], Peruzzi, Nelson José [UNESP], La Scala, Newton [UNESP], Panosso, Alan Rodrigo [UNESP]
Tipo de documento: Artigo
Idioma: eng
Título da fonte: Repositório Institucional da UNESP
Texto Completo: http://dx.doi.org/10.1007/s11356-023-26824-6
http://hdl.handle.net/11449/248684
Resumo: 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 (R2 adj): 0.70 and root mean square error (RMSE): 1.02 µmol m−2 s−1], RP (R2 adj: 0.48 and RMSE: 1.07 µmol m−2 s−1) and GS (R2 adj: 0.70 and RMSE: 1.05 µmol 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. Graphical abstract: [Figure not available: see fulltext.]
id UNSP_ba0a160b7956118189d7b705c67feafc
oai_identifier_str oai:repositorio.unesp.br:11449/248684
network_acronym_str UNSP
network_name_str Repositório Institucional da UNESP
repository_id_str 2946
spelling Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian CerradoClimate changeEnvironmental modelingSoil respirationTropical ecosystemsSoil 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 (R2 adj): 0.70 and root mean square error (RMSE): 1.02 µmol m−2 s−1], RP (R2 adj: 0.48 and RMSE: 1.07 µmol m−2 s−1) and GS (R2 adj: 0.70 and RMSE: 1.05 µmol 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. Graphical abstract: [Figure not available: see fulltext.]Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)Department Engineering and Exact Sciences School of Agricultural and Veterinarian Sciences São Paulo State University (FCAV/UNESP), Via de Acesso Prof. Paulo Donato Castellane S/N, São PauloDepartment of Phytotecnics Faculty of Engineer (FEIS/UNESP), Avenida Brasil – Centro, São PauloDepartment Engineering and Exact Sciences School of Agricultural and Veterinarian Sciences São Paulo State University (FCAV/UNESP), Via de Acesso Prof. Paulo Donato Castellane S/N, São PauloDepartment of Phytotecnics Faculty of Engineer (FEIS/UNESP), Avenida Brasil – Centro, São PauloCAPES: 001Universidade Estadual Paulista (UNESP)Canteral, Kleve Freddy Ferreira [UNESP]Vicentini, Maria Elisa [UNESP]de Lucena, Wanderson Benerval [UNESP]de Moraes, Mário Luiz Teixeira [UNESP]Montanari, Rafael [UNESP]Ferraudo, Antonio Sergio [UNESP]Peruzzi, Nelson José [UNESP]La Scala, Newton [UNESP]Panosso, Alan Rodrigo [UNESP]2023-07-29T13:50:48Z2023-07-29T13:50:48Z2023-05-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/article61052-61071http://dx.doi.org/10.1007/s11356-023-26824-6Environmental Science and Pollution Research, v. 30, n. 21, p. 61052-61071, 2023.1614-74990944-1344http://hdl.handle.net/11449/24868410.1007/s11356-023-26824-62-s2.0-85152426212Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengEnvironmental Science and Pollution Researchinfo:eu-repo/semantics/openAccess2024-06-06T13:43:59Zoai:repositorio.unesp.br:11449/248684Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T23:23:52.708776Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false
dc.title.none.fl_str_mv Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
title Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
spellingShingle Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
Canteral, Kleve Freddy Ferreira [UNESP]
Climate change
Environmental modeling
Soil respiration
Tropical ecosystems
title_short Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
title_full Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
title_fullStr Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
title_full_unstemmed Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
title_sort Machine learning for prediction of soil CO2 emission in tropical forests in the Brazilian Cerrado
author Canteral, Kleve Freddy Ferreira [UNESP]
author_facet Canteral, Kleve Freddy Ferreira [UNESP]
Vicentini, Maria Elisa [UNESP]
de Lucena, Wanderson Benerval [UNESP]
de Moraes, Mário Luiz Teixeira [UNESP]
Montanari, Rafael [UNESP]
Ferraudo, Antonio Sergio [UNESP]
Peruzzi, Nelson José [UNESP]
La Scala, Newton [UNESP]
Panosso, Alan Rodrigo [UNESP]
author_role author
author2 Vicentini, Maria Elisa [UNESP]
de Lucena, Wanderson Benerval [UNESP]
de Moraes, Mário Luiz Teixeira [UNESP]
Montanari, Rafael [UNESP]
Ferraudo, Antonio Sergio [UNESP]
Peruzzi, Nelson José [UNESP]
La Scala, Newton [UNESP]
Panosso, Alan Rodrigo [UNESP]
author2_role author
author
author
author
author
author
author
author
dc.contributor.none.fl_str_mv Universidade Estadual Paulista (UNESP)
dc.contributor.author.fl_str_mv Canteral, Kleve Freddy Ferreira [UNESP]
Vicentini, Maria Elisa [UNESP]
de Lucena, Wanderson Benerval [UNESP]
de Moraes, Mário Luiz Teixeira [UNESP]
Montanari, Rafael [UNESP]
Ferraudo, Antonio Sergio [UNESP]
Peruzzi, Nelson José [UNESP]
La Scala, Newton [UNESP]
Panosso, Alan Rodrigo [UNESP]
dc.subject.por.fl_str_mv Climate change
Environmental modeling
Soil respiration
Tropical ecosystems
topic Climate change
Environmental modeling
Soil respiration
Tropical ecosystems
description 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 (R2 adj): 0.70 and root mean square error (RMSE): 1.02 µmol m−2 s−1], RP (R2 adj: 0.48 and RMSE: 1.07 µmol m−2 s−1) and GS (R2 adj: 0.70 and RMSE: 1.05 µmol 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. Graphical abstract: [Figure not available: see fulltext.]
publishDate 2023
dc.date.none.fl_str_mv 2023-07-29T13:50:48Z
2023-07-29T13:50:48Z
2023-05-01
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv http://dx.doi.org/10.1007/s11356-023-26824-6
Environmental Science and Pollution Research, v. 30, n. 21, p. 61052-61071, 2023.
1614-7499
0944-1344
http://hdl.handle.net/11449/248684
10.1007/s11356-023-26824-6
2-s2.0-85152426212
url http://dx.doi.org/10.1007/s11356-023-26824-6
http://hdl.handle.net/11449/248684
identifier_str_mv Environmental Science and Pollution Research, v. 30, n. 21, p. 61052-61071, 2023.
1614-7499
0944-1344
10.1007/s11356-023-26824-6
2-s2.0-85152426212
dc.language.iso.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv Environmental Science and Pollution Research
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv 61052-61071
dc.source.none.fl_str_mv Scopus
reponame:Repositório Institucional da UNESP
instname:Universidade Estadual Paulista (UNESP)
instacron:UNESP
instname_str Universidade Estadual Paulista (UNESP)
instacron_str UNESP
institution UNESP
reponame_str Repositório Institucional da UNESP
collection Repositório Institucional da UNESP
repository.name.fl_str_mv Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)
repository.mail.fl_str_mv
_version_ 1808129516591120384