资源环境科技发展态势分析平台

This study developed a stand table projection system for interior Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco var. glauca (Beissn.) Franco) in British Columbia, Canada. Simulation data, obtained by running Prognosis(BC) (with 150 year projections) using input derived from 140 permanent sample plots, were utilized for model construction. First, a whole-stand attributes (number of trees per ha, quadratic mean diameter, and arithmetic mean diameter) prediction model was constructed. Weibull parameters for the diameter distribution of the future stand were estimated from the predicted stand attributes using two parameter prediction methods (via seemingly unrelated regression and the cumulative distribution function) and one parameter recovery method. The future stand table was projected with the estimated Weibull parameters and compared with two stand diameter distribution adjustment algorithms, yielding a total of nine parameter estimation-diameter adjustment combinations. The outputs from the stand table projection model were evaluated with repeated measures data covering 20 years from 18 independent plots, as well as compared with the outcomes predicted by PrognosisBc for those plots. The diameter class adjustment procedure applying tree mortality prior to growth, combined with the cumulative distribution parameter prediction method, performed best among the nine combinations we compared. The stand table projection system yielded similar results to Prognosis(BC) on the evaluation data and is expected to be a useful tool for managing forests sustainably and effectively.