Modelling Tree Height-Diameter Relationship in a Mixed Forest Plantation

Abstract

The diameters and heights of the trees are two of the most important measurements in a forest inventory for biomass estimation and sustainable management. Measuring tree height in a forest stand is time consuming and costly, it is necessary to develop models that accurately estimate tree heights from easily measured variables (tree diameter). This study aims to develop models for estimating tree height in a forest plantation located in North-central, Nigeria. The systematic sampling method was used to twenty-one 0.09 ha sample plots in study area. Data on tree height and diameter were collected.  Artificial neural network (ANN) model, support vector regression (SVR) model, and four empirical nonlinear models were tested for estimating tree height. The models were evaluated using the Coefficient of Determination, Residual standard Error, Mean Bias and Akaike’s Information Criterion. The results showed that the SVR model best predicted tree heights in the study area than the ANN and empirical nonlinear models. The SVR model explained about 94% variance associating with the dependent variable. The SVR model can be conveniently used for predicting the height of trees in the study area.