An individual tree growth modeling system was developed for the loblolly pine (Pinus taeda L.) - shortleaf pine (Pinus echinata Mill.) forest type in Louisiana using USDA Forest Service Forest Inventory and Analysis (FIA) data. In this study, the loblolly pine - shortleaf pine forest type was divided into three species groups: loblolly pine, other pines, and hardwoods. The growth system includes models for individual tree survival, diameter growth, height growth, and change in crown ratio for each of the three species groups. A multivariate extension of a two-step, model-based interpolation method is proposed to estimate parameters of annual tree growth equations based on measurements from 7-year growth periods. Results based on evaluation statistics such as mean difference, mean absolute difference, and fit index show that the two-step interpolation method is clearly superior to the averaging method, which is a typical linear interpolation method. The two-step method produced growth equations that project tree and plot attributes more accurately than the averaging method. Additionally, a Monte Carlo simulation analysis reveals that parameter estimates obtained from the two-step method are closer to the true parameter values than those from the averaging method. The approach described here should be useful for estimating parameters of other systems of annual tree growth equations from periodic measurements. Models such as the one presented here may be useful for projecting the characteristics of undisturbed forest inventory plots, e.g., FIA plots, and providing more up-to-date inventory estimates.
All Science Journal Classification (ASJC) codes
- Global and Planetary Change