Use este identificador para citar ou linkar para este item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/346550
Tipo: Artigo
Título: Using temporal stability to estimate soya bean yield: a case study in Paraná state, Brazil
Autor(es): Figueiredo, Gleyce K. D. A.
Brunsell, Nathaniel A.
Rocha, Jansle V.
Lamparelli, Rubens A. C.
Picoli, Michelle C. A.
Resumo: Crop identification is an important task in the process of yield estimation; however, sometimes it can be difficult when using images of medium to low spatial resolution due to mixing of heterogeneous areas within the pixel. Therefore, selecting pixels that best represent an area/crop could be an alternative to input data in yield estimate models. The objective of this study was to select soya bean pixels based on temporal stability technique and test the ability of these pixels to predict yield. The study was conducted at county level in Paraná state, Brazil, during 11 years of soya bean growing season. To estimate yield, we created a linear regression model and used accumulated enhanced vegetation index during four periods according to soya bean phenological stages, which are as follows: emergence to maturity, emergence to flowering, flowering to grain filling, and flowering to maturity. Among all periods of the crop season, emergence to flowering showed the lowest precision while flowering to maturity was the period with the best agreements when compared with official data; the root mean square error ranged from 0.07 to 0.37 t ha–1. The temporal stability method has proven to be an efficient tool to select pixel that could represent the crop for predicting yield. In addition, we could verify the most suitable period for making soya bean yield prediction.
Palavras-chave: Estabilidade temporal
Soja
País: Reino Unido
Editor: Taylor & Francis
Tipo de Acesso: Fechado
Identificador DOI: 10.1080/01431161.2016.1148280
Endereço : https://www.tandfonline.com/doi/full/10.1080/01431161.2016.1148280
Data do documento: 2016
Aparece nas coleções:FEAGRI - Artigos e Outros Documentos

Arquivos associados a este item:
Não existem arquivos associados a este item.


Os itens no repositório estão protegidos por copyright, com todos os direitos reservados, salvo quando é indicado o contrário.