本文用世界上几个地区较为系统的测量数据进一步检验了有关ＬＭＳＳ树林遮蔽衰减的ＥＲＳ预测模式，发现该模式有较大的预测误差。也对Ｌｏｏ模式进行了评述和分析。然后，基于所述测量数据，通过回归分析，对ＥＲＳ和Ｌｏｏ模式进行了改进，其中考虑了仰角和光学遮蔽度两者的影响。所得改进模式都足够简单易用，而且预测精度较高。 In this paper, we further test the ERS prediction model of forest shelter attenuation in LMSS with more systematic measurements in several regions of the world and find that this model has larger prediction errors. The Loo model is also reviewed and analyzed. Then, based on the measurement data, the ERS and Loo modes were improved by regression analysis, taking into account both the elevation angle and optical shadowing. The resulting improvements are simple enough to use and have high predictive accuracy.