Statistical Method for Detecting Structural Change in the Growth Process
Statistical Method for Detecting Structural Change in the Growth Process
复制标题
DOI：
10.1111/j.1541-0420.2007.00844.x
复制
发表时间：
2008-03
期刊：
影响因子：
1.9
通讯作者：
Yoshiyuki Ninomiya;A. Yoshimoto
中科院分区：
文献类型：
--
作者：
Yoshiyuki Ninomiya;A. Yoshimoto
由于个别树木之间的竞争和其他改变生长环境的外源因素，每棵树都遵循自己的生长趋势，随着时间的推移，生长中会发生一些结构性变化。在本文中，提出了一种新的方法来检测在生长过程中的结构变化。我们制定的方法作为一个简单的信号检测的统计测试，而不构建任何特定的模型的结构变化。为了评估测试的p值，由于规则分布理论的不足，开发了管法。利用日本柳杉人工林的两组树径生长数据，分析了间伐对生长过程的影响，认为间伐是一种结构性变化。我们的结果表明，所提出的方法是有用的，以识别减薄引起的结构变化。我们还提供了测试的大小和功率方面的方法的属性。
Summary Due to competition among individual trees and other exogenous factors that change the growth environment, each tree grows following its own growth trend with some structural changes in growth over time. In the present article, a new method is proposed to detect a structural change in the growth process. We formulate the method as a simple statistical test for signal detection without constructing any specific model for the structural change. To evaluate the p‐value of the test, the tube method is developed because the regular distribution theory is insufficient. Using two sets of tree diameter growth data sampled from planted forest stands of Cryptomeria japonica in Japan, we conduct an analysis of identifying the effect of thinning on the growth process as a structural change. Our results demonstrate that the proposed method is useful to identify the structural change caused by thinning. We also provide the properties of the method in terms of the size and power of the test.