VARIABILIDADE ESPACIAL DA PRODUTIVIDADE DA SOJA E DOS ATRIBUTOS DO SOLO: DIAGNÓSTICO DE INFLUÊNCIA LOCAL UTILIZANDO O MODELO GEOESTATÍSTICO WAVE
Ano de defesa: | 2024 |
---|---|
Autor(a) principal: | |
Orientador(a): | |
Banca de defesa: | , , , |
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: | |
Palavras-chave em Inglês: | |
Área do conhecimento CNPq: | |
Link de acesso: | https://tede.unioeste.br/handle/tede/7462 |
Resumo: | Soybeans constitute Brazil's primary agribusiness export and rank among the world’s most significant commodities. This study highlights soybean productivity and the growing need for advanced precision agriculture (PA) techniques to maximize economic returns while minimizing environmental impact. Geostatistics, as a core statistical approach in PA, enables accurate interpolations for mapping soybean productivity and soil chemical and physical attributes, which are critical for informed agricultural decision-making. The Wave geostatistical model is emphasized for its unique capacity to depict spatial dependence, particularly when semivariance suggests the presence of a hole effect. In this research, novel diagnostic techniques for local influence on spatial data were developed and tested specifically with the Wave model, allowing the identification of influential observations that may distort semivariance and compromise spatial model accuracy. These techniques refined spatial parameter estimates and enhanced the quality of interpolated maps, which are essential for delineating more precise management zones, optimizing input applications, and boosting economic returns (Article 1). Results demonstrated that the Wave model effectively captures complex spatial dependence patterns, notably those arising from heterogeneous soils commonly found in agricultural settings. Additionally, combining Thin Plate Spline (TPS) interpolation with kriging with external drift proved to be a robust approach to improving estimates in unsampled areas by leveraging data from fixed covariates (Article 2). Through these techniques, producers can identify zones with high productive potential and others needing targeted intervention. This approach is vital for optimizing input usage, reducing waste, lessening environmental impact, and enhancing economic profitability. |