论文部分内容阅读
基于276株实测生物量数据,构建了东北林区红松、臭冷杉、红皮云杉和兴安落叶松4个天然针叶树种总量及各分项生物量一元、二元可加性生物量模型.采用似然分析法判断总量及各分项生物量异速生长模型的误差结构(可加型或相乘型),而模型参数估计采用非线性似乎不相关回归模型方法.结果表明:经似然分析法判断,4个天然树种总量及各分项生物量异速生长模型的误差结构都是相乘型的,对数转换的可加性生物量可以被选用.各树种可加性生物量模型的调整后确定系数Ra2为0.85~0.99,平均相对误差为-7.7%~5.5%,平均相对误差绝对值<30.5%.增加树高可以显著提高各树种可加性生物量模型的拟合效果和预测能力,而且总量、地上和树干生物量模型效果较好,树根、树枝、树叶和树冠生物量模型效果较差.所建立的可加性生物量模型的预测精度为77.0%~99.7%(平均92.3%),可以很好地预估东北林区天然红松、臭冷杉、红皮云杉和兴安落叶松的生物量.
Based on the measured biomass data of 276 plants, four natural coniferous species of Pinus koraiensis, Sophora alopecuroides, Picea koraiensis and Larix gmelinii in northeastern China were constructed, and the biomass of one-component and binary adducible biomass The error structure (additive or multiplicative) of allometric and allometric biomass growth model was judged by the likelihood analysis method, while the model parameter estimation was based on the non-linear regression model which seemed unrelated.The results showed that: Likelihood analysis shows that the error structures of the total amount of 4 natural species and the allometric growth model of each sub-species are all multiplicative types, and the additive biomass of logarithmic transformation can be selected. After adjusting the biomass model, the coefficient of determination Ra2 was 0.85 ~ 0.99, the average relative error was -7.7% ~ 5.5%, the average relative error was less than 30.5% .Increasing the tree height could significantly improve the model of additive tree biomass Combined effect and predictive ability, and the total amount, the aboveground and trunk biomass model effect is better, the biomass model of the root, branch, leaves and crown is less effective.The prediction accuracy of the additive biomass model established is 77.0% ~ 99.7% (average 92.3%), can be very To forecast natural pine, Abies, red spruce and larch Larix biomass Northeast Forest.