Definição de zonas de manejo utilizando análises multivariadas

• Ano de defesa: 2023
• Autor(a) principal: Jean Lucas Pereira Oliveira
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Cana-de-açúcar; Agricultura de precisão; Variabilidade espacial; Irrigação
• Link de acesso: https://hdl.handle.net/11449/254757

Specific management using a management zone is used on agricultural properties with the aim of maximizing agricultural yield and the application of inputs. However, it is known that to create ZM, soil, climatic and crop variables can be taken into account throughout the development cycle. Therefore, the objective of this project was to delineate management zones using multivariate analysis and the fuzzy c-means algorithm. The experiment was conducted in the Jaboticabal region, São Paulo. To compose the multivariate analyses, chemical and physical parameters of the soil, meteorological variables, vegetation indices and productivity were evaluated. Yield data were obtained from the sampling grid. Meteorological variables were obtained by an automated meteorological station. To obtain vegetation indices, satellite images from the Planet Scope platform were used. After laboratory analysis, the data was subjected to descriptive statistics and correlation analysis was performed to select related attributes to define the management zones. Then, the spatial variability of these attributes was evaluated and spatial distribution maps were created using geostatistical tools. After that, the fuzzy c-means algorithm was performed to delineate ZMs using the MZA software.