Study on the R-Based Dynamic Construction Techniques for Stand Density Management Diagrams

With an investigation into the procedure and techniques involved in constructing a stand density management diagram (SDMD) and in light of users’ real needs, this study designs a system for the dynamic construction of SDMDs based on R language, for the purpose of overcoming the complexities of model fitting in producing a diagram, the difficulties in constructing a diagram and the high demands for users’ expertise. The interactive modelling technique between C# language and R language is applied to the research, with which different sets of model fitting results are compared and optimal results are chosen. The whole procedure, from data collection, model fitting (including outlier data elimination, prototype model fitting and model testing) and diagram construction, is streamlined and automated. Continuous forest inventory data for the Larix principis-rupprechtii plantation in China’s Inner Mongolia is used and two sets of prototype fitting results of the SDMD are compared. With equivalent height curves as the example, this paper makes the goodness of fit of the binomial theory of density effect (R²=0.9231) better than that of the power theory of density effect (R²=0.8354). The model with the best goodness of fit is finally selected for the dynamic construction of a SDMD. Application of this generated SDMD indicates a relatively high accuracy and practical value.