文献类型：期刊文献

中文题名：随机分布的角尺度置信区间及其应用

英文题名：Uniform Angle Index( W) Confidence Interval of the Random Distribution and Its Application

作者：胡艳波[1] 惠刚盈[1] 王宏翔[1] 李远发[1] 赵中华[1] 刘文桢[2]

第一作者：胡艳波

机构：[1]中国林业科学研究院林业研究所,国家林业局林木培育重点实验室;[2]甘肃省小陇山林业试验局

年份：2014

卷号：27

期号：3

起止页码：302-308

中文期刊名：林业科学研究

外文期刊名：Forest Research

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：国家自然科学基金"基于相邻木关系的混交林树种分布格局测度方法研究"(31370638)

语种：中文

中文关键词：林分结构;空间结构参数;角尺度;置信区间;格局检验方法

外文关键词：stand structure; spatial structure parameter; uniform angle index; confidence interval; patterntesting method

分类号：S718.54

摘要：角尺度通过描述相邻木围绕参照树的均匀性来进行林木水平分布格局的判定。研究分析80 000个模拟随机分布林分的角尺度均值的标准差(σ珚W)发现:随机分布林分的角尺度均值的标准差主要受模拟株数(N)的影响,模拟窗口的大小对其的影响可以忽略不计;模拟株数与角尺度均值的标准差关系极为紧密,表现为模拟株数越少,标准差越大,且标准差随着模拟株数的增加而变小,这种关系用幂函数σ珚W=0.210 34N-0.488 72能够很好地表达,相关指数R2高达0.998;研究基于统计学正态分布原理建立了随机分布林分角尺度均值置信区间,95%的置信限为0.5±1.96σ珚W=0.5±1.96×0.210 34N-0.488 72;99%的置信限为0.5±2.58σ珚W=0.5±2.58×0.210 34N-0.488 72;并有当林分或某一种群的林木分布的角尺度均值珚W在所建立的置信区间[珚WLd,珚WLu]内,即在相应的置信水平上判断为随机,若珚W>珚WLu时为团状分布,珚W<珚WLd时为均匀分布。这个与调查株数有关的判定标准的提出进一步完善了角尺度理论,并为林分空间结构参数角尺度在实践中的应用提供了简洁方法。
The Uniform angle index（W） method is a parameter analyzing forest spatial structure through describing the evenness of neighboring trees around reference tree to judge the distribution pattern of trees. 80 000 simulated randomly distributed stands were analyzed, and it was found that the standard deviation of the mean value of the W （W） of those stands was mainly influenced by the tree number of simulation （N）, while the simulation window size was not so important for the result. The tree number of simulation related closely to the standard deviation of W, the fewer the former are, the larger the latter will be. Moreover, the standard deviation decreased with the tree number of simulation. Such relationship can be well expressed by the power function σW = 0. 210 34N^-0.48872 , and the correlation index RE was up to 0. 998. Based on the normal distribution principle in statistics, a confidence interval was established for the Wof the Random Distribution stands, and the 95% confidence limit is 0.5 ± 1.96σW = 0. 5 ±1.96 × 0. 210 34N^-0.48872 , while the 99% confidence limit is0. 5 ±2.58σW = 0. 5 ±2.58 × 0. 210 34N^-0.48872 When the W of a stand or a specific population is within the established confidence interval [ WLd, WLu ] , then it should be described as a random distribution on relevant confidence level, and clustered distribution should be judged if W 〉WLu , while uniform distribution if W 〈 WLd.