本文是用九块森林小区摸拟抽样,对已知全部收获重量和估测样区重量相比较。根据样区重量与断面积的相关关系估测其平均值,即平均重量:(∑样木重量)/(∑样木断面积)×(样区断面积)估测结果得出很小的负偏差,并发现与用对数回归估测方程求出的可变量很接近。用断面积比率法估测的最小值稍差些,但其平均值和最大值始终比用对数回归方法要好些。使用五个径阶的分层随机抽样仅比随机抽样好些。 In this paper, nine forest plots are used to simulate sampling, comparing the weight of all the harvested crops with that of the estimated plots. According to the relationship between the sample weight and the cross-sectional area, the average value is estimated, that is, the average weight: (Σ sample weight) / (Σ sample area) × (sample area) Deviations and found to be very close to the variables found using the log-regression estimation equation. The minimum estimated by the area-to-area ratio method is slightly worse, but the mean and maximum values ​​are always better than using the logarithmic regression method. Stratified random sampling using five diameter scales is only better than random sampling.