林分材積推定の新しい方法(II) : 林分材積推定量の母分散と本法の精度

概要

論文の詳細を見る
The author proposed a new method of estimating stand volume by point sampling in the previous paper. The aim of this paper is to examine the population variance of this estimator of stand volume, and to compare the relative variance, as a measurement of the precision of the method with that of BITTERLICH's method from the analytical viewpoint, and also with that of BITTERLICH's method and that of KITAMURA's critical height method, through computer simulation. The results follow. The population variance of this estimator is directly proportional to BAF and H, where H is a height that is greater than the maximum tree height in the stand. Therefore, the value of H needs to be established at a height that is just a little greater than the maximum tree height in the stand, and a small BAF needs to be employed in order to increase the precision. Next, if the population variance of this method and that of BITTERLICH's method do not include the covariance, the ratio of relative variance of this method to that of BITTERLICH'S method is approximately equivalent to H/F・h, where F is a stand form factor, and h is mean height. In practice, this ratio, through computer simulation, was 9.7〜18.3. Consequently, the approximate equation (H/F・h) may be valid in this sense, which gives an index of the actual forest stand.
日本森林学会の論文
1980-08-25