关键词: 樟子松人工林, 哑变量, 胸高断面积, 碳储量生长模型, 联立方程组

Abstract: To scientifically predict the dynamics of carbon storage in Pinus sylvestris var. mongolica plantation, we identified the Weibull model as the optimal model for describing the variation of stand carbon storage with stand age, based on data from fixed monitoring plots of P. sylvestris var. mongolica plantation in Heilongjiang Province. Following a reparameterization approach, we further constructed a generalized model C1 incorporating site class index (SCI) and stand density index (SDI), and a generalized model C2 incorporating SCI and stand basal area (BAS). To quantify regional effects, we introduced a regional dummy variable into model C1, resulting in model C3. Then, we established a simultaneous equation system for carbon storage and BAS using nonlinear seemingly unrelated regression, forming the carbon storage model systems. The results showed that the optimized models C1, C2, and C3 had coefficients of determination (R²) of 0.9856, 0.9968, and 0.9862, respectively, with root mean square errors (RMSE) of all models being less than 3 t·hm^-2, indicating high model stability and predictive accuracy. A comparison between models C1 and C2 showed that BAS had a greater influence on carbon storage estimation than SDI. Model C3, based on regional dummy variables, categorized 26 forest areas into three regions. Under the same stand mean age, SCI, and SDI conditions, carbon storage followed an order of Region 2 (Wandashan area) > Region 1 (Xiaoxing’anling area and Zhangguangcailing area) > Region 3 (plain area), confirming regional vari-ations in carbon accumulation. In the simultaneous model systems, all sub-models achieved R² values greater than 0.98 and relative root mean square errors (rRMSE) below 9%, demonstrating strong generalizability and stability of the established system. The independent models (C1, C2 and C3) and the model system developed here have their specific focuses and are suitable for precise prediction of stand carbon storage in different practical application scenarios, providing a scientific basis for carbon sink assessment and management decision-making in P. sylvestris var. mongolica plantations.

Key words: Pinus sylvestris var. mongolica plantation, dummy variable, stand basal area, stand carbon storage growth model, simultaneous equations