Complementarity of proximal and orbital sensing techniques for exploring soil variability

• Grant number: 24/13229-0
• Support Opportunities: Scholarships in Brazil - Master
• Effective date (Start): December 01, 2024
• Effective date (End): March 31, 2026
• Field of knowledge: Agronomical Sciences - Agricultural Engineering
• Principal Investigator: Lucas Rios do Amaral
• Grantee: João Vítor Fiolo Pozzuto
• Host Institution: Faculdade de Engenharia Agrícola (FEAGRI). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 22/03160-8 - Soil spatial variability mapping and optimized sampling supported by sensing techniques: bases for a more efficient and sustainable precision agriculture, AP.PNGP.PI

Abstract

The use of sensors for mapping the spatial variability of agricultural lands is already consolidated through various sensors in precision agriculture, however, the complementarity/redundancy of these is little studied. Therefore, this work aims to verify whether different soil sensing techniques are complementary, studied both from an unsupervised perspective (only the sensors) and supervised, for soil physical-chemical attributes and grain productivity. Three agricultural areas in the state of São Paulo will be used, in the municipalities of Caiuá (200ha - 2samples/ha), Cosmópolis (103ha - 6samples/ha), and Sales Oliveira (72ha - 6samples/ha), where soil fertility samples were collected, apparent electrical conductivity (ECa), magnetic susceptibility (SM), VIS-NIR and XRF spectra and bands of a synthetic image of exposed soil (SYSI). Complementarity/redundancy will be studied through collinearity (Spearman's linear correlation, VIF, and CI), grouping by PCA, stepwise multiple linear regression (RLM) and PLSR, as well as nonlinear regression, by Random Forest (RF).