通过应用机载激光雷达数据，在分析云南省中部的78块样地的基础上提出2个预测森林不同生物特性的统计模型(加权平均高度的预测模型和生物量的预测模型)，并讨论了预测结果及其精确性。从激光雷达数据中提取了2组变量(树冠高度变量组和植被密度变量组)作为自变量，采用逐步回归方法进行自变量选择。结果表明，激光雷达数据与森林的平均树高和地上各部分生物量有很强的相关性。对于3种不同森林类型(针叶林，阔叶林和混交林)，平均树高估测均能达到比较高的精度；生物量的估测结果是针叶林优于阔叶林，混交林的生物量与激光雷达数据则没有明显相关性。最后，对回归分析的结果和影响预测精度的因素进行讨论,认为预测结果的精度可能与森林类型、激光雷达采样时间和采样密度以及坐标误差等因素有关。

Light Detection and Ranging (LiDAR) is one of the most promising technologies in forestry, which shows potential for timely and accurate measurements of forest biophysical properties over time. This study explores several regression models relating variables derived from airborne laser scanner for the estimation of various forest metrics, and discusses the results of prediction concluding accuracy. These prediction models use 78 plots with radius of 7.5 m or 15 m in Kunming, Yunnan province, China. Two series of variables are provided from the airborne laser scanner data, one is canopy height and to the other canopy density. These variables are used as independent variables in the regressions. The stepwise regression analysis has been used to select various independent variables. The results show high correlation between forest metrics and variables derived from airborne laser scanner. For the three different forest types (coniferous, broad-leaf and mixed), all the prediction of mean heights are accurate. However, for the predictions of above ground biomass, the result of coniferous is better than broad-leaf, while there is no signifi cant correlation between the biomass of mixed and the laser variables. Finally, the results of regression and factors affect the accuracy of prediction are discussed. The accuracy of prediction may be relate to the forest type, sampling time and density of laser scanning and position errors.