MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES

Detalhes bibliográficos
Autor(a) principal: de Azevedo Silva, Vinícius
Data de Publicação: 2023
Outros Autores: Mateus Peixoto, Santos, Francisco Lledo
Tipo de documento: Artigo
Idioma: por
eng
Título da fonte: Holos
Texto Completo: http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340
Resumo: Precipitation forecasting is essential for sectors such as water resources management and urban planning. In this study, a deep learning model was developed to predict rainfall in Brazilian cities, focusing on the municipality of Barra Mansa, Rio de Janeiro. Four neural network architectures were tested: FCN, Resnet, ResCNN and InceptionTime. Among them, FCN stood out significantly, presenting the lowest error rates and the best overall adjustment. The study highlights the ability of deep learning, especially through the FCN (Fully Convolutional Network - Segmented) architecture, to make accurate predictions and uncover hidden rainfall patterns. Such discoveries have great potential to improve rainfall forecasting systems and assist in decision-making in areas that require accurate climate information.
id IFRN-3_255e6a60d6756b23ef602a8908fc67f0
oai_identifier_str oai:holos.ifrn.edu.br:article/16340
network_acronym_str IFRN-3
network_name_str Holos
repository_id_str
spelling MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUESPREVISÃO DE PRECIPITAÇÃO MENSAL NO MUNICÍPIO DE BARRA MANSA/RJ USANDO TÉCNICAS DE DEEP LEARNING TIME SERIESForecasting, precipitation, rainfall, deep learning, neural networksPrevisão, precipitação, chuvas, apredizagem profunda, redes neurais.Precipitation forecasting is essential for sectors such as water resources management and urban planning. In this study, a deep learning model was developed to predict rainfall in Brazilian cities, focusing on the municipality of Barra Mansa, Rio de Janeiro. Four neural network architectures were tested: FCN, Resnet, ResCNN and InceptionTime. Among them, FCN stood out significantly, presenting the lowest error rates and the best overall adjustment. The study highlights the ability of deep learning, especially through the FCN (Fully Convolutional Network - Segmented) architecture, to make accurate predictions and uncover hidden rainfall patterns. Such discoveries have great potential to improve rainfall forecasting systems and assist in decision-making in areas that require accurate climate information.A previsão de precipitações é essencial para setores como gestão de recursos hídricos e planejamento urbano. Neste estudo, foi desenvolvido um modelo de aprendizagem profunda (deep learning) para prever chuvas em cidades brasileiras, com foco no município de Barra Mansa, Rio de Janeiro. Foram testadas quatro arquiteturas de redes neurais: FCN, Resnet, ResCNN e InceptionTime. Dentre elas, a FCN se destacou significativamente, apresentando os menores índices de erro e o melhor ajuste global. O estudo evidencia a capacidade da aprendizagem profunda, especialmente através da arquitetura FCN, em fazer previsões precisas e desvendar padrões ocultos das chuvas. Tais descobertas possuem grande potencial para aprimorar sistemas de previsão de chuvas e auxiliar na tomada de decisões em áreas que necessitam de informações climáticas acuradas.  Instituto Federal de Educação, Ciência e Tecnologia do Rio Grande do Norte2023-12-18info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionTextoinfo:eu-repo/semantics/otherapplication/pdfapplication/pdfhttp://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/1634010.15628/holos.2023.16340HOLOS; v. 5 n. 39 (2023): v.5 (2023)1807-1600reponame:Holosinstname:Instituto Federal do Rio Grande do Norte (IFRN)instacron:IFRNporenghttp://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340/3846http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340/3847Barra Mansa, 1940 - 2023Barra Mansa, 1940 - 2023https://creativecommons.org/licenses/by-nc-nd/4.0info:eu-repo/semantics/openAccessde Azevedo Silva, ViníciusMateus PeixotoSantos, Francisco Lledo2023-12-28T03:14:36Zoai:holos.ifrn.edu.br:article/16340Revistahttp://www2.ifrn.edu.br/ojs/index.php/HOLOSPUBhttp://www2.ifrn.edu.br/ojs/index.php/HOLOS/oaiholos@ifrn.edu.br||jyp.leite@ifrn.edu.br||propi@ifrn.edu.br1807-16001518-1634opendoar:2023-12-28T03:14:36Holos - Instituto Federal do Rio Grande do Norte (IFRN)false
dc.title.none.fl_str_mv MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
PREVISÃO DE PRECIPITAÇÃO MENSAL NO MUNICÍPIO DE BARRA MANSA/RJ USANDO TÉCNICAS DE DEEP LEARNING TIME SERIES
title MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
spellingShingle MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
de Azevedo Silva, Vinícius
Forecasting, precipitation, rainfall, deep learning, neural networks
Previsão, precipitação, chuvas, apredizagem profunda, redes neurais.
title_short MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
title_full MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
title_fullStr MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
title_full_unstemmed MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
title_sort MONTHLY RAINFALL FORECAST IN THE MUNICIPALITY OF BARRA MANSA/RJ USING DEEP LEARNING TIME SERIES TECHNIQUES
author de Azevedo Silva, Vinícius
author_facet de Azevedo Silva, Vinícius
Mateus Peixoto
Santos, Francisco Lledo
author_role author
author2 Mateus Peixoto
Santos, Francisco Lledo
author2_role author
author
dc.contributor.author.fl_str_mv de Azevedo Silva, Vinícius
Mateus Peixoto
Santos, Francisco Lledo
dc.subject.por.fl_str_mv Forecasting, precipitation, rainfall, deep learning, neural networks
Previsão, precipitação, chuvas, apredizagem profunda, redes neurais.
topic Forecasting, precipitation, rainfall, deep learning, neural networks
Previsão, precipitação, chuvas, apredizagem profunda, redes neurais.
description Precipitation forecasting is essential for sectors such as water resources management and urban planning. In this study, a deep learning model was developed to predict rainfall in Brazilian cities, focusing on the municipality of Barra Mansa, Rio de Janeiro. Four neural network architectures were tested: FCN, Resnet, ResCNN and InceptionTime. Among them, FCN stood out significantly, presenting the lowest error rates and the best overall adjustment. The study highlights the ability of deep learning, especially through the FCN (Fully Convolutional Network - Segmented) architecture, to make accurate predictions and uncover hidden rainfall patterns. Such discoveries have great potential to improve rainfall forecasting systems and assist in decision-making in areas that require accurate climate information.
publishDate 2023
dc.date.none.fl_str_mv 2023-12-18
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
Texto
info:eu-repo/semantics/other
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340
10.15628/holos.2023.16340
url http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340
identifier_str_mv 10.15628/holos.2023.16340
dc.language.iso.fl_str_mv por
eng
language por
eng
dc.relation.none.fl_str_mv http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340/3846
http://www2.ifrn.edu.br/ojs/index.php/HOLOS/article/view/16340/3847
dc.rights.driver.fl_str_mv https://creativecommons.org/licenses/by-nc-nd/4.0
info:eu-repo/semantics/openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-nd/4.0
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.coverage.none.fl_str_mv Barra Mansa, 1940 - 2023
Barra Mansa, 1940 - 2023
dc.publisher.none.fl_str_mv Instituto Federal de Educação, Ciência e Tecnologia do Rio Grande do Norte
publisher.none.fl_str_mv Instituto Federal de Educação, Ciência e Tecnologia do Rio Grande do Norte
dc.source.none.fl_str_mv HOLOS; v. 5 n. 39 (2023): v.5 (2023)
1807-1600
reponame:Holos
instname:Instituto Federal do Rio Grande do Norte (IFRN)
instacron:IFRN
instname_str Instituto Federal do Rio Grande do Norte (IFRN)
instacron_str IFRN
institution IFRN
reponame_str Holos
collection Holos
repository.name.fl_str_mv Holos - Instituto Federal do Rio Grande do Norte (IFRN)
repository.mail.fl_str_mv holos@ifrn.edu.br||jyp.leite@ifrn.edu.br||propi@ifrn.edu.br
_version_ 1798951616764706816

Registros relacionados