Illustration of the Varying Coefficient Model for a Tree Growth Analysis from the Age and Space Perspectives

Abstract

One reason why trees grow differently over the year may be due to their relative spatial condition. In this paper, we introduce a statistical model called the varying coefficient model to estimate effects on the tree growth from the age and relative space perspectives. This model is a superior way to understand the tree growth, which is because 1) the space effects can be added in the growth curve model and visually checked by drawing contours and cubic diagrams, and 2) both the space and the age effects are estimated with their confidence interval values, and also the hypothesis test can be performed, which improve the confidence of the estimation result. The model is applied to the data of Cryptomeria japonica, the longitude and the latitude are used as the space information, and the significant spatial effect of the DBH on the stem volume was evident from the estimation result.