论文部分内容阅读
基于黑龙江省带岭林业局大青川林场80株人工兴安落叶松解析木数据和Logistic生长模型,分别考虑单木效应和样地效应,利用S-PLUS软件中的NLME过程拟合非线性材积生长模型,采用赤池信息准则(AIC)、贝叶斯信息准则(BIC)、对数似然值和似然比检验等模型评价指标对不同模型的精度进行比较.结果表明:当考虑单木效应影响时,b1、b2、b3(分别代表Logistic模型中的渐进、尺度和形状的随机参数)同时作为随机参数时模型拟合效果最好;当考虑样地效应影响时,b1作为随机参数时模型拟合效果最好.基于单木效应和样地效应的混合模型的拟合精度高于基本模型(Logistic生长模型),考虑单木效应影响的混合模型的精度高于考虑样地效应影响的模型.模型检验结果表明,随机效应模型不但能反映单木材积的总体平均变化趋势,还能反映个体之间的差异;随机效应模型通过校正随机参数值能提高模型的预测精度.
Based on 80 trees of Larix chinensis var. Larch and Logistic growth model in Daqingchuan Forest Farm, Heilongjiang Province, the single-wood effect and plots effect were considered respectively. The nonlinear volume growth model was fitted by NLME process in S-PLUS software (AIC), Bayesian Information Criterion (BIC), log-likelihood ratio and likelihood ratio test were used to compare the accuracy of different models.The results showed that when considering the effect of single wood effect , b1, b2, b3 (which represent the random parameters of gradual, scale and shape respectively in Logistic model) are the best parameters for model fitting at the same time as random parameters. When considering the effect of plot effect, The best fitting accuracy was obtained for the mixed model based on the single-wood effect and the plots effect, and the accuracy of the mixed model considering the single-wood effect was higher than that of the model considering the effect of the plots. The test results show that the random effects model can not only reflect the overall average trend of the single timber volume, but also reflect the differences among individuals. The random effects model can correct the random parameters Can improve the prediction accuracy of the model.