Quantitative pedology: VIS-NIR-SWIR spectrometry and digital soil mapping

Author(s)

Leonardo Ramírez López

• Document type: Master's Dissertation
• Press: Piracicaba.
• Institution: Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC)
• Defense date: 2009-06-17

Examining board members

Jose Alexandre Melo Dematte; Maria de Lourdes Mendonça Santos Brefin; Jose Paulo Molin

• Advisor: Jose Alexandre Melo Dematte
• Field of knowledge: Agronomical Sciences - Agronomy
• Indexed in: Banco de Dados Bibliográficos da USP-DEDALUS; Biblioteca Digital de Teses e Dissertações - USP
• Location: Universidade de São Paulo. Biblioteca Central da Escola Superior de Agricultura Luiz de Queiroz; t631.47 94525; R173p

Abstract

The routine soil analysis is traditionally used on the evaluation of soil attributes related to land use potential, and the assessment of fertility. It is costly and time consuming, making it inaccessible for small farmers, and hampering the applicability of precision agriculture on soil management in Brazil. Currently, pedometrics is providing the possibility of incorporating in soil science sophisticated techniques that can help to reduce the cost of obtaining information and improve the understanding about how several soil processes works. One of the more recent topics on pedometrics is the soil reflectance spectroscopy. Through the soil reflected energy is possible to infer several soil properties, although optimum accuracy levels in the spectral estimation of soil attributes have not yet reached. In this sense, the aim of this study was basically answer the following questions: a. The exchangeable bases have specific spectral bands or the spectral responses of theses depends on the clay mineral?; b. the calibration of models by using only some specific spectral bands may improve the prediction performance?; c. What is the influence of the accuracy of prediction models on maps constructed with predicted soil attributes?; D. How calibration set size affect the accuracy of the models?; e. How the calibration of models for prediction of soil attributes related to soil weathering may assist the digital soil mapping? (AU)