SciELO - Scientific Electronic Library Online

 
vol.21 issue1STAND-LEVEL PROGNOSIS OF EUCALYPTUS CLONES USING ARTIFICIAL NEURAL NETWORKSTREE DIAMETER GROWTH FOLLOWING SILVICULTURAL TREATMENTS IN A SEMI-DECIDUOUS SECONDARY FOREST IN CENTRAL BRAZIL author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


CERNE

Print version ISSN 0104-7760

Abstract

MENDONCA, Adriano Ribeiro de; CARVALHO, Samuel de Pádua Chaves e  and  CALEGARIO, Natalino. GENERALIZED MIXED HEIGHT-DIAMETER MODELS FOR Eucalyptus sp. HEIGHT PREDICTION. CERNE [online]. 2015, vol.21, n.1, pp.107-115. ISSN 0104-7760.  http://dx.doi.org/10.1590/01047760201521011191.

The purpose of this study was to evaluate the theory of the generalized mixed models and the modeling of the heteroscedasticity of variance in the fit of two nonlinear models for the height-diameter relationship in Eucalyptus sp. Data from sample trees of different genotypes (clone and seed) and ages were used. According to the analysis, the asymptotic model achieved the best results compared to the sigmoid model. The accuracy of the models increased using the theory of mixed effects with the addition of covariates and modeling of the heteroscedasticity of variance.

Keywords : mixed effect; probability models; heteroscedasticity; covariates.

        · abstract in Portuguese     · text in Portuguese     · Portuguese ( pdf )