Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil

SUMMARY: The increasing awareness of global climate change has drawn attention to the role of forests as mitigators of this process as they act as carbon sinks to the atmosphere. Understanding the process of carbon storage in forests and its drivers, as well as presenting consistent models for their...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Dantas,Daniel, de Castro Nunes Santos Terra,Marcela, Baldissera Schorr,Luis Paulo, Calegario,Natalino
Lenguaje:	English
Publicado:	Universidad Austral de Chile, Facultad de Ciencias Forestales 2021
Materias:	artificial intelligence artificial neural networks support vector machines forest biomass
Acceso en línea:	http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0717-92002021000100131
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:scielo:S0717-92002021000100131
record_format	dspace
spelling	oai:scielo:S0717-920020210001001312021-07-25Machine learning for carbon stock prediction in a tropical forest in Southeastern BrazilDantas,Danielde Castro Nunes Santos Terra,MarcelaBaldissera Schorr,Luis PauloCalegario,Natalino artificial intelligence artificial neural networks support vector machines forest biomass SUMMARY: The increasing awareness of global climate change has drawn attention to the role of forests as mitigators of this process as they act as carbon sinks to the atmosphere. Understanding the process of carbon storage in forests and its drivers, as well as presenting consistent models for their estimation, is a current demand. In this sense, the aim of this study was to evaluate the performance of machine learning techniques: support vector machines (SVM) and to propose a new nonlinear model extracted from the training of an artificial neural network (ANN) in the modeling of above ground carbon stock in a secondary semideciduous forest. SVM and ANN construction and training process considered independent variables selected by stepwise: minimum DBH (diameter of breast height - 1.3 m), maximum DBH, mean DBH, total height and number of trees, all by plot. SVM and the model extracted from ANN were applied to the data set intended for validation. Both techniques presented satisfactory performance in modeling carbon stock by plot, with homogeneous distribution and low dispersion of residues and predicted values close to those observed. Analysis criteria indicated superior performance of the model extracted from the artificial neural network, which presented a mean relative error of 6.94 %, while the support vector machine presented 13.52 %, combined with lower bias values and higher correlation between predictions and observations.info:eu-repo/semantics/openAccessUniversidad Austral de Chile, Facultad de Ciencias ForestalesBosque (Valdivia) v.42 n.1 20212021-01-01text/htmlhttp://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0717-92002021000100131en10.4067/S0717-92002021000100131
institution	Scielo Chile
collection	Scielo Chile
language	English
topic	artificial intelligence artificial neural networks support vector machines forest biomass
spellingShingle	artificial intelligence artificial neural networks support vector machines forest biomass Dantas,Daniel de Castro Nunes Santos Terra,Marcela Baldissera Schorr,Luis Paulo Calegario,Natalino Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
description	SUMMARY: The increasing awareness of global climate change has drawn attention to the role of forests as mitigators of this process as they act as carbon sinks to the atmosphere. Understanding the process of carbon storage in forests and its drivers, as well as presenting consistent models for their estimation, is a current demand. In this sense, the aim of this study was to evaluate the performance of machine learning techniques: support vector machines (SVM) and to propose a new nonlinear model extracted from the training of an artificial neural network (ANN) in the modeling of above ground carbon stock in a secondary semideciduous forest. SVM and ANN construction and training process considered independent variables selected by stepwise: minimum DBH (diameter of breast height - 1.3 m), maximum DBH, mean DBH, total height and number of trees, all by plot. SVM and the model extracted from ANN were applied to the data set intended for validation. Both techniques presented satisfactory performance in modeling carbon stock by plot, with homogeneous distribution and low dispersion of residues and predicted values close to those observed. Analysis criteria indicated superior performance of the model extracted from the artificial neural network, which presented a mean relative error of 6.94 %, while the support vector machine presented 13.52 %, combined with lower bias values and higher correlation between predictions and observations.
author	Dantas,Daniel de Castro Nunes Santos Terra,Marcela Baldissera Schorr,Luis Paulo Calegario,Natalino
author_facet	Dantas,Daniel de Castro Nunes Santos Terra,Marcela Baldissera Schorr,Luis Paulo Calegario,Natalino
author_sort	Dantas,Daniel
title	Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
title_short	Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
title_full	Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
title_fullStr	Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
title_full_unstemmed	Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil
title_sort	machine learning for carbon stock prediction in a tropical forest in southeastern brazil
publisher	Universidad Austral de Chile, Facultad de Ciencias Forestales
publishDate	2021
url	http://www.scielo.cl/scielo.php?script=sci_arttext&pid=S0717-92002021000100131
work_keys_str_mv	AT dantasdaniel machinelearningforcarbonstockpredictioninatropicalforestinsoutheasternbrazil AT decastronunessantosterramarcela machinelearningforcarbonstockpredictioninatropicalforestinsoutheasternbrazil AT baldisseraschorrluispaulo machinelearningforcarbonstockpredictioninatropicalforestinsoutheasternbrazil AT calegarionatalino machinelearningforcarbonstockpredictioninatropicalforestinsoutheasternbrazil
_version_	1718444261044649984

Machine learning for carbon stock prediction in a tropical forest in Southeastern Brazil

Ejemplares similares