【目的】 论证基于易测林木因子模拟林下辐射方法体系中,尺度化加权直射荫蔽度与加权进界邻体直射荫蔽度两种林下直射辐射成分测度指数的模拟准确性,寻求基于简单、易测林木因子较为合理、准确估算林下任意点位直射辐射的模型结构特征。【方法】基于阔叶红松(Pinus koraiensis)林林分调查信息,建立林分体素模型模拟阔叶红松林下直射透射率。将模拟的阔叶红松林下直射透射率作为验证标准,分别以爬山算法对尺度化加权直射荫蔽度与加权进界邻体直射荫蔽度中的各项参数进行寻优;以林下直射透射率与参数寻优后的尺度化加权直射荫蔽度、加权进界邻体直射荫蔽度间的Pearson、Spearman相关系数及线性回归决定系数为指标,评价尺度化加权直射荫蔽度、加权进界邻体直射荫蔽度对林下直射辐射的模拟性能。【结果】尺度化加权直射荫蔽度在最小邻体林木选取尺度(r_min)与最大邻体林木选取尺度(r_max)分别取6.332、13.609 m时,与阔叶红松林下直射透射率间具有较高的线性回归拟合精度,此时其对应的校正系数为10.440,与林下直射透射率间的Pearson、Spearman相关系数及线性回归决定系数分别为-0.581、-0.645及0.338。加权进界邻体直射荫蔽度在最大邻体林木选取临界值(T_max)与最小邻体林木选取临界值(T_min)分别为1.965与0.502时,与阔叶红松林下直射透射率间的线性回归拟合精度最高,此时其对应的校正系数为17.465,与林下太阳直射透射率间的Pearson、Spearman相关系数及线性回归决定系数分别为-0.738、-0.695与0.545。【结论】尺度化加权直射荫蔽度与加权进界邻体直射荫蔽度均能较好地反映林下太阳直射辐射,其中以加权进界邻体直射荫蔽度表现效果更优。在构建基于林木因子的林下直射辐射模拟模型时,应以合理地选取潜在影响林下直射辐射的邻体林木作为重点,同时应寻求能合理表达不同时刻林下直射辐射状态的简单方式。

【Objective】 This study aims to assess the simulation accuracy of two measures of understory beam radiation: the scaled weighted beam shading degree and the weighted inside-boundary neighbor beam shading degree. These measures are part of a forest understory radiation simulation system based on easy measurable tree factors. The study also seeks to identify model structural characteristics using these simple, measurable tree factors to provide a reasonable and accurate estimate of beam radiation at any understory location. 【Method】A stand pixel model was established using broadleaf Pinus koraiensis (Korean pine) forest stand survey information and it simulated the understory beam transmittance. This transmittance served as a benchmark to evaluate the scaled weighted beam shading degree and the weighted inside-boundary neighbor beam shading degree. Various parameters for these degrees were optimized using a hill-climbing algorithm. The simulation performance of these measures was assessed by comparing the proportions of beam radiation transmitted through gaps, using Pearson and Spearman correlation coefficients and linear regression determination coefficients between understory beam transmittance and the two degrees under the optimized parameters.【Result】The results indicate that the scaled weighted beam shading degree provided the best linear regression fit for understory beam transmittance in the broadleaf Korean pine forest, with a neighbor tree selection minimum scale (r_min) of 6.332 m and a maximum scale (r_max) of 13.609 m. The Pearson and Spearman correlation coefficients, as well as the linear regression determination coefficient between this measure and understory beam transmittance, were -0.581, -0.645 and 0.338, respectively. By contrast, the weighted inside-boundary neighbor beam shading degree showed a linear regression fit for understory beam transmittance, with a neighbor tree selection maximum threshold (T_max) of 1.965 and a minimum threshold (T_min) of 0.502. The Pearson and Spearman correlation coefficients, as well as the linear regression determination coefficient for this measure, were -0.738, -0.695, and 0.545, respectively. 【Conclusion】(1) Both the scaled weighted beam shading degree and the weighted inside-boundary neighbor beam shading degree effectively reflect the understory beam radiation, with the latter performing better. (2) When constructing a simulation model of beam radiation in forests based on tree factors, it is crucial to select neighboring trees that influence beam radiation and to seek a straightforward method to accurately represent beam radiation at various times.