A linkage among whole-stand model, individual-tree model and diameter-distribution model

Abstract

Stand growth and yield models include whole-stand models, individual-tree models and diameter-distribution models. In this study, the three models were linked by forecast combination and parameter recovery methods one after another. Individual-tree models combine with whole-stand models through forecast combination. Forecast combination method combines information from different models, disperses errors generated from different models, and then improves forecast accuracy. And then the forecast combination model was linked to diameter-distribution models via parameter recovery methods. During the moment estimation, two methods were used, arithmetic mean diameter and quadratic mean diameter method (A-Q method), and arithmetic mean diameter and diameter variance method (A-V method). Results showed that the forecast combination for predicting stand variables outperformed over the stand-level and tree-level models respectively; A-V method was superior to A-Q method on estimating Weibull parameters; these three different models could be linked very well via forecast combination and parameter recovery.