Abstract:

Based on the sample data of 1102 pine trees in the distribution area of Pinus taiwanensis in Henan Province,which were divided into 776 modeling samples and 300 test samples according to the diameter order distribution,this paper used the least square nonlinear regression modeling method,so as to conduct parameter estimation on the 8 alternative binary standing volume models,6 univariate standing volume models,and 7 ground diameter volume models. Indexes like the sum of squares of deviation,correlation index,total relative error,average value of relative error,average value of absolute value of relative error and prediction accuracy were used for model optimization,the applicability test was carried out by using the test samples through three test indexes: total relative error,average value of absolute value of relative error and F test. The results showed that the optimal model of binary standing volume of Pinus taiwanensiswas variable parameter Yamamoto model;the optimal model of one-dimensional standing volume was Desakoku Meyer model;the optimal model of ground diameter standing volume was determined. The establishment of stand volume model of Pinus taiwanensis and its application in tabulation made up for the defect of no stand volume table in the distribution area of Pinus taiwanensis in Henan Province,and can effectively improve the accuracy and efficiency of investigation and research of Pinus taiwanensis.

Key words: Pinus taiwanensis, volume model, parameter estimation, suitability test