欢迎访问《植物研究》杂志官方网站,今天是 分享到:

植物研究 ›› 2015, Vol. 35 ›› Issue (6): 929-936.doi: 10.7525/j.issn.1673-5102.2015.06.022

• 论文 • 上一篇    下一篇

大兴安岭天然落叶松林相容性立木含碳量模型研究

蒋蕾;刘兆刚*;董灵波;孙云霞   

  1. 东北林业大学林学院,哈尔滨 150040
  • 出版日期:2015-11-20 发布日期:2016-01-18
  • 基金资助:
     

Carbon Stocks Model of Compatible Individual Tree in the Natural Larch Forest of Daxing’an Mountains

JIANG Lei;LIU Zhao-Gang*;DONG Ling-Bo;SUN Yun-Xia   

  1. Northeast Forestry University,Harbin 150040
  • Online:2015-11-20 Published:2016-01-18
  • Supported by:
     

摘要: 森林是生态系统的重要组成部分,是改善全球气候变暖趋势的机体,森林的固碳能力越来越受到重视,而研究森林中立木的含碳量对森林生态系统量化固碳能力具有重要意义。本文基于大兴安岭44株天然落叶松解析木实测数据和各器官碳密度样木数据,借鉴相容性生物量模型的思想来研究相容性立木含碳量模型,基于基础模型y=aDby=a(D2H)b,利用非线性度量误差模型系统构建了总含碳量和干、枝、叶、根四个分含碳量之间相容的一元和二元立木含碳量模型。通过比较各模型的拟合优度和独立检验统计量,计算拟合优度结果显示:在一元和二元基础模型下的4个器官,树干的确定系数R2分别为0.960,0.985,都是2个模型中各器官确定系数的最高值,树枝、树叶和树根相对偏低,但均达到了85%以上,说明建立的模型可行;模型检验统计量表明:一元和二元基础模型的树干模拟效率EF值为0.904,0.951,相应的预估精度P值为80.5%,85.5%,其次是树枝模拟效率0.830,0.898,精度都在70%以上,树叶和树根预估精度偏低,其值在70%左右。综上研究结果表明:二元立木含碳量预估模型的拟合及预测精度优于一元模型。

关键词: 大兴安岭, 天然落叶松, 非线性度量误差, 相容性含碳量模型

Abstract: Forest is an important part of the ecosystem and is the body in improving the global warming trend with the ability of the forest carbon sequestration and more attention, and the trees of forest carbon stocks in forest ecosystems is important to quantify carbon sequestration significance. With 44 wood analytic data of Daxing’an Mountains natural larch and various organs carbon sample wood density data, we drew the ideological compatibility biomass model to study compatibility carbon storage model. By the models of y=aDb and y=a(D2H)b, we used nonlinear equations of measurement error to establish the compatible of single and binary models of total carbon stocks and stem, branches, leaves and roots. By comparing the goodness of fit and independent test statistic of the model. For the goodness of fitting, the determination coefficient R2 of the trunk in the single and binary models of four organs were 0.960 and 0.985, respectively, and all two values were the maximum. But for the branches, leaves and roots, the values were relatively low, reaching more than 85%, indicating that the overall model were feasible. By the model test, the efficiencies of EF values in different analog trunk model were 0.904 and 0.951, with the estimated accuracy of 80.5% and 85.5%, respectively, followed by the branches simulation efficiencies of 0.830 and 0.898 with the accuracy of more than 70%, and the leaf and root forecast accuracies were low with the values of 70%. The forecast binary model accuracy of fitting and prediction was better than that in the single model.

Key words: Daxing’an Mountains, natural larch, nonlinear measurement errors model, compatible carbon stocks model

中图分类号: