Avaliação de modelos de regressão não linear na cinética de liberação de potássio de resíduos orgânicos / Evaluation nonlinear regression models on the potassuim kinetic release from organic residues
AUTOR(ES)
Walmes Marques Zeviani
DATA DE PUBLICAÇÃO
2009
RESUMO
Potassium (K) is required in great amounts by crops, but its availability in Brazilian soils is less than its demand. Its supply can be efficiently made from organic sources when we know the pattern of release. Nonlinear models are appropriate in these situations since then estimate quantities of practical interest, and they have goodness of fit. Although its inferential process is based on asymptotic arguments there are ways to know the nonlinearity intensity. In this work we evaluate the nonlinearity, through the curvatures of Bates &Watts, bias of Box and the least squares estimator sampling properties by simulation study of two nonlinear regression models. These models estimate K readily releasable content, their halflife time release and the release rate of K slow release. The data were obtained from the study over time of the K release from 11 organic waste in combination with 3 soils. The Exponential model gave more precise parameters estimates than the Ratio and closer relations between the asymptotic and simulation results. The Exponential model was more appropriate in terms of inferential and practical application, since by all measures showed lower non-linearity.
ASSUNTO(S)
half-life time tempo de meia vida curvature of bates e watts bootstrap estatistica curvatura de bates e watts vício de box bias of box bootstrap
