Dinâmica espectro-temporal da soja e do feijão por meio de sensoriamento remoto terrestre

• Ano de defesa: 2017
• Autor(a) principal: Prudente, Victor Hugo Rohden
• Orientador(a): Mercante, Erivelto
• Banca de defesa: Maggi, Marcio Furlan , Souza, Carlos Henrique Wachholz de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Paraná; Cascavel
• Programa de Pós-Graduação: Programa de Pós-Graduação em Engenharia Agrícola
• Departamento: Centro de Ciências Exatas e Tecnológicas
• País: Brasil
• Palavras-chave em Português: Estimativas; Índices de vegetação; Parâmetros biofísicos
• Palavras-chave em Inglês: Estimates; Vegetation indexes; Biophysical parameters
• Área do conhecimento CNPq: CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA
• Link de acesso: http://tede.unioeste.br/handle/tede/3062

The spatio-temporal dynamics of agricultural crops can be monitored through Vegetation Indices (VI) on multiple dates. The use of IV in a given period of crop development or throughout its cycle, has generated good relations with agronomic parameters. There are several IV, the NDVI (Normalized Difference Vegetation Index) being one of the most used. NDVI correlates with some agronomic variables such as plant height, leaf area index (LAI), dry matter, chlorophyll index (CI), soil cover degree, plant height and photosynthetically active intercepted radiation (PAIR). Hence, the objective of this work is to compare and intercalibrate NDVI spectral-temporal profiles obtained by different terrestrial remote Sensors, and to use these profiles to estimate biophysical parameters (LAI, CI, PAIR, and Plant Height) for soybean and bean, in addition to estimating and studying soybean yield through models with spectral-temporal data obtained by different terrestrial remote sensors. The ground sensors used are the GreenSeeker 505 Handheld active sensor and the passive hyperspectral FieldSpec4 Standard-Res model. The NDVI values of the passive (NDVI FS) and active (NDVI GS) sensors presented high correlation (soybean rs = 0.90 and bean rs = 0.93) and the linear models respond well to an intercalibration considering the entire phenological cycle of soybean (R² = 0.83, RMSE = 0.0161 e dr = 0.85) and beans (R² = 0.92, RMSE = 0.0038 e dr = 0.89). The saturation of the NDVI values at the survey dates that are located during the vegetative peak of the crops and the presence of adverse climatic events decreased the precision of the models in certain periods of the phenological cycle of the crops. The variables LAI and Height for both crops and CI for the soybean crop obtained better results when estimated by the values of NDVI GS. PAIR for both crops and CI for bean obtained better adjustments considering the models using NDVI FS values. For the variable productivity, there was no significant correlation with the NDVI GS and NDVI FS. However, it presented relations with the accumulation of reflectance observed in the RED and NIR bands for both sensors as a result of different collection dates. The best relations of productivity were with the accumulation of reflectance values of the spectroradiometer bands on DAS76 and DAS71 dates.