Estimation of the height of forest species in natural regeneration using artificial neural networks

Authors

  • Emanuel França Araújo Universidade Federal do Espirito Santo
  • Milton Marques Fernandes Universidade Federal de Sergipe
  • Jeferson Pereira Martins Silva Universidade Federal do Espirito Santo
  • Sustanis Horn Kunz Universidade Federal do Espirito Santo
  • Marcia Rodrigues de Moura Fernandes Secretaria de Estado do Desenvolvimento Urbano e Sustentabilidade

DOI:

https://doi.org/10.18316/rca.v14i3.6142

Keywords:

Atlantic Forest, Secondary Succession, Artificial Intelligence.

Abstract

The use of neural networks has been used in various branches of science. In vegetation studies, neural networks have been used mainly to estimate tree biomass, height and diameter of tree individuals, but there are few studies with natural regeneration. In this context, the objective was to estimate, through artificial neural networks, the height (H) of regenerating species in an area of a Semideciduous Ombrophilous Forest fragment. The database came from individuals of tree species from regeneration of 20 plots of 10 m2 (2 x 5 m), in São Cristóvão, SE. To estimate total H by artificial neural networks, Multilayer Perceptron networks were tested. RNA was efficient in the estimation of H of the regenerative stratum in an area of a Semideciduous Ombrophilous Forest fragment. RNA 12 with 14 neurons was more efficient to estimate the height of the most abundant species in the area. The greater number of individuals with lower heights promoted an overestimation of the height.

Author Biographies

Emanuel França Araújo, Universidade Federal do Espirito Santo

Doutorando em Ciências Florestais da UFES

Milton Marques Fernandes, Universidade Federal de Sergipe

Professor Doutor do Departamento de Ciências Florestais na Área de Ecologia Florestal

Jeferson Pereira Martins Silva, Universidade Federal do Espirito Santo

Doutorando em Ciências Florestais da UFES

Sustanis Horn Kunz, Universidade Federal do Espirito Santo

Professora Doutora do Departamento de Engenharia Florestal e Engenharia Industrial na Àrea de Restauração Ecológica

Marcia Rodrigues de Moura Fernandes, Secretaria de Estado do Desenvolvimento Urbano e Sustentabilidade

Doutora em Ciências Florestais pela UFES

Downloads

Published

2020-12-17

Issue

Vol. 14 No. 3 (2020)

Section

Artigos

License

Authors must submit their manuscripts to be published in this journal agree with the following terms:

  1. Authors maintain the copy rights and concede to the journal the right of first publication, with the paper simultaneously licensed under the License Creative Commons attribution that permits the sharing of the paper with recognition of authorship and initial publication in this journal.

  2. Since the articles are presented in this journal of public access, they are of free use, with their own attributions for educational and non-commercial purposes.

 

Licença Creative Commons

The RCA Journal - REVISTA DE CIÊNCIAS AMBIENTAIS in: http://www.revistas.unilasalle.edu.br/index.php/Rbca was licensed with a Creative Commons License Creative Commons - Attribution - Noncommercial 3.0 Not Adapted.