Espectrorradiometria em cultivo da soja Glycine max (L.) Merr. durante ciclo vegetativo

• Ano de defesa: 2011
• Autor(a) principal: Felipe, João Paulo de Mello
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/26339/0013000006x1v
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Geociências; UFSM; Programa de Pós-Graduação em Geomática
• Palavras-chave em Português: Espectrorradiômetro agricultura de precisão; Sensoriamento remoto; Soja; Glycine max; Estimativa de produção; Spectrorradiometer; Precision agriculture; Remote sensing; Soya; Estimated production; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS
• Link de acesso: http://repositorio.ufsm.br/handle/1/9589

Soy is one of the products of most relevance to the Brazilian economy. Estimating soybean productivity through remote sensing is a potential tool for precision farming, qualifying and quantifying the productive potential of crops. The main objective of the work was to relate the data obtained through field from radiometric dates with the productivity of soybean cultivation and validate the data obtained through remote sensing platforms orbital (CBERS and LANDSAT) with the use of vegetation index. The study area is located at the Federal University of Santa Maria, with a total area of 16.14 hectares. Readings were made in each of the 15 points of working with the grid Espectrorradiometer. With the field data and Satellite images of vegetation indices were calculated. In 2009/2010 the best multiple regression models found to have been for the groups of vegetation Indices 1 (CRI, Near-Infraredt B4, REP VARI and WBI), 4 (CRI, REP, NDMI, VARI and SAVI) and 11 (Red B3, SAVI, REP and VARI) where the coefficients of determination and determination adjusted reached 97.70% and values 96.40%; 98.00% and 96.30% and 97.72% and 96.41% for groups 1, 4 and 11 respectively, and have low values of standard deviation. Showing that the combination of vegetation index of the groups in question can be used to estimate crop with good accuracy. It is important to highlight that all groups had good correlations with soybean productivity with 43 days after planting. The multiple regression analysis and Stepwise Backward with the vegetation Indices calculated with data from LANDSAT images of dates 24/01, 09/14/02 and 04, 2010, did not show significant values for any regressions.