Comparative analysis of nonlinear height-dbh models for Quercus species in South Korea

This study was conducted to analyze the tree height-diameter at breast height (dbh) relationship of the five Quercus species in South Korea using the six commonly used nonlinear growth models. The data used in this study was acquired from the 5(th) National Forest Inventory (NFI) in South Korea collected in the year 2007 and 2008. In this inventory, the dbh of all trees in the established permanent plots were measured while 10 representative sampled trees were chosen for their height. This led to the development of the height-dbh equation which can determine the unknown height of a tree based on its dbh. A total of 8,364 individual tree total height and dbh measurements or 80% of the dataset were used in the initial equation development while the 20% or 2,091 individual trees were randomly selected and used to validate the equations developed for each species. On the other hand, the combined data (100%) were used for final model fitting or estimating the parameters. Results showed that the coefficient of determination (R-2) of all six models in all five species ranged from 0.947 to 0.954 while the root mean square error (RMSE) ranged from 2.159 to 2.532. On the other hand, the overall mean difference (MD) ranged from -0.004 to 0.024 m while the absolute mean difference (AMD) ranged from 1.662 to 1.982 m. All the six models were able to determine the relationships of height and dbh and fit the data well but the Exponential followed by Korf/Lundqvist were the best models based on the four performance criteria using the rank analysis. The equation developed will be very significant for forest resource managers in acquiring accurate timber inventory information while reducing the cost and time.