基于星载LiDAR波形数据的森林胸高断面积估测研究

Forest Basal Area Estimation Based on Spaceborne LiDAR Waveform Data

为探索星载LiDAR在森林胸高断面积上的应用潜力,利用星载激光雷达ICESat/GLAS波形数据对吉林汪清林区的森林胸高断面积进行估测。通过对ICESat/GLAS波形数据波形进行处理,提取一系列波形特征参数,利用回归估计方法分析波形特征参数与胸高断面积的相关关系,建立预测模型对森林胸高断面积进行估测。结果表明:在所提波形特征参数中,波形长度、H5和H10与胸高断面积在三次函数拟合时具有明显的相关性,判定系数R2分别为0.813、0.737和0.613,预测精度P分别为0.951、0.932和0.901,综合分析最终确定波形长度估测模型最好,该模型可稳定准确地估测出胸高断面积。表明星载LiDAR是未来估测森林胸高断面积一种有效的方法,并为进一步的森林蓄积量研究提供了一种技术手段。

In order to explore the potential application of spaceborne LiDAR in the estimation of the basal area at breast height（BA）of forests,an investigation was carried out in Wangqing forest Farm of Jilin Province by using spaceborne lidar ICESat/GLAS waveform data.The data were processed to extract a series of waveform characteristic parameters.Correlations between obtained parameters and BA were analyzed by using regression estimation method,and the prediction models were established to estimate BA.Among the extracted characteristic parameters of the waveform,significant correlations were observed between BA and the parameters,such as waveform length,H5,and H10,fitted by cubic function estimation method.The corresponding R2 values were 0.813,0.737and 0.613,respectively,and prediction accuracy Pvalues were 0.951,0.932and 0.901,respectively.After comprehensive analysis,it was concluded that waveform length based model was the best one for estimating BA.The results indicated that it would be an effective method to apply spaceborne LiDAR in estimating BA,and an technical measure for the further estimation of forest volume.