%0 Conference Proceedings
%4 sid.inpe.br/marte2/2017/10.27.14.06.34
%2 sid.inpe.br/marte2/2017/10.27.14.06.35
%@isbn 978-85-17-00088-1
%F 59841
%T Modelagem digital de frações granulométricas do solo na região da Cuesta de Botucatu - SP: aplicação do NDVI como variável auxiliar
%D 2017
%A Carvalho, Tania Maria de,
%A Nicolete, Donizeti Aparecido Pastori,
%A Zimback, Célia Regina Lopes,
%@electronicmailaddress dnicolete@gmail.com
%E Gherardi, Douglas Francisco Marcolino,
%E Aragão, Luiz Eduardo Oliveira e Cruz de,
%B Simpósio Brasileiro de Sensoriamento Remoto, 18 (SBSR)
%C Santos
%8 28-31 maio 2017
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%P 5281-5288
%S Anais
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%X Many pedological information are required in land use planning, management of agroforestry activities and environmental studies, usually as a soil map. Currently, the soil attributes mapping applies quantitative modeling, and explanatory covariates representing the factors of soil forming equation. Legacy soil data were used at this modelling process and the Normalized Difference Vegetation Index (NDVI) is one of the most common radiometric indexes used as predictors for mapping soil size fractions. The aim of this work was to examine the potential of the NDVI for predicting sand and clay fractions of the soils, in an area where the vegetation are in recomposition process, using an hybrid model of digital soil modeling. The NDVI of two periods (coincident with soil sampling and current) along with terrain attributes were applied as auxiliary variables predictors of grain size fractions at two depths, using as target variable soil attributes data from a semi - detailed survey of soils. The regression-kriging technique (RK) was applied, using a multiple linear regression (RLM) and posterior sum with a kriging map of the residuals to obtain a prediction map. The values of the coefficient or determination were low, suggesting poor performance of the models. The results showed that the slope and profile curvature were the most significant variables in the prediction process and the NDVI coinciding with the soil sampling time was more important, especially for the subsurface layer.
%9 Modelagem espacial
%@language pt
%3 59841.pdf