Abstract: Traditional forest growth and yield models have been criticized for their inability to provide precise spatial information for forest and ecosystem management. In this study the spatial patterns of trees within 50 spruce-fir plots in the Northeast, USA were modeled by a Gibbs point process model with three pair potential functions. In general, 82%-84% of these 50 plots were modeled well by the Gibbs model. However, the complete spatial random (CSR) and regular spatial patterns were modeled better than the clustered plots. Further, empirical regression models were developed to predict the two parameters of the Gibbs model using the available stand variables as predictors such as stand density, basal area, mean tree diameter, mean tree height, mean crown length, and mean crown width. The simulation results showed that 81% of the 50 plots were satisfactorily predicted by the empirical regression models. Among them, 100% of the CSR plots, 71% of the regular plots, and 56% of the clustered plots were predicted well by the empirical regression models. Three example plots were selected to illustrate the similarity between simulated tree locations and observed ones.

Key words: spatial point pattern, spatial point process, Gibbs model, Ripley's K-function, Markov Chain Monte Carlo (MCMC), Metropolis algorithm, Takacs-Fiksel method