基于随机效应的兴安落叶松材积生长模拟

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.

Simulation of Larix gmelinii tree volume growth based on random effect

Based on the stem analysis data of 80 sample trees in dahurian larch(Larix gmelinii) plantations of Daqingchuan Forest Farm,Dailing Forest Bureau in Heilongjiang Province and the Logistic growth model,the NLME procedure of S-PLUS software was adopted to fit the nonlinear tree volume growth models,with consideration of individual tree effect and plot effect,and the evaluation statistics such as AIC,BIC,Log Likelihood,and likelihood ratio test were used to compare the prediction precisions of the models.The r...