我们试验研究了利用多级遥感图象信息通过多阶不等概抽样在云南省西双版纳自治州进行森林资源清查。试验所用遥感图象有陆地卫星图象和小比例尺及大比例尺航空象片。抽样设计采用概率与单元的大小成比例的抽样。试验包括下列几个主要组成部份。 1 利用陆地卫星图象进行一阶单元有林地面积成数预估值的判读。 2 随机不等概抽取一阶样本单元。 3 利用航空象片进行二阶单元森林蓄 We experimentally studied the forest inventory in the Xishuangbanna Autonomous Prefecture of Yunnan Province by using multilevel unequal sampling using multi-level remote sensing image information. Remote sensing images used in the experiments included terrestrial satellite images and small-scale and large-scale aerial photographs. Sampling design uses a sample whose probability is proportional to the size of the cell. The test consists of the following major components. 1 Use land satellite imagery to interpret the first-order unit estimated forestland area percentages. 2 randomly sampled first order sample unit. 3 aerial photographs for second-order unit forest storage