论文部分内容阅读
基于遥感影像精确估算区域人工林蓄积量对于人工林的管理具有重要意义。本研究以北半球面积最大的华北落叶松(Larix principis-rupprechtii)人工林林场——塞罕坝机械林场为研究区,采用逐步回归方法分别建立了华北落叶松人工林地面实测蓄积量与L波段ALOS PALSAR雷达数据、Landsat-8 OLI数据、以及共用ALOS PALSAR数据与Landsat-8OLI数据之间的不同模型,提出了一种基于雷达影像准确估算人工林蓄积量的方法。通过模型精度验证,找到落叶松人工林蓄积量估算最佳模型,并据此获得研究区不同林龄华北落叶松人工林蓄积量空间分布。结果表明:在建立的4个回归模型中,ALOS PALSAR雷达数据HV极化模式归一化后向散射系数与华北落叶松人工林蓄积量指数形式的回归模型预测精度最高(R2=0.67,相对RMSE=26.78%,P<0.01);据此模型估算获得研究区2010年华北落叶松人工林总蓄积量为4.7×106m3及其相应的空间分布图;当单位蓄积量达到250~300 m3·hm-2时,ALOS PALSAR雷达数据HV极化模式归一化后向散射系数对蓄积量的变化不敏感,达到了饱和;林龄<20 a、20~30 a、30~40 a和>40 a的华北落叶松人工林单位蓄积量分别为78、97、136和127 m3·hm-2。本研究提出了一种基于雷达影像准确估算人工林蓄积量的方法。
Accurate estimation of the accumulation of regional plantations based on remote sensing images is of great importance to the management of plantations. In this study, Larix principis-rupprechtii plantation forest farm-Saihanba mechanical forest farm with the largest area in the northern hemisphere was used as the study area. The stepwise regression method was used to establish the relationship between the measured ground stock volume and L-band ALOS PALSAR radar data, Landsat-8 OLI data and different models of common ALOS PALSAR data and Landsat-8OLI data, a method of accurately estimating the stock volume of plantations based on radar images was proposed. Through the model accuracy verification, the optimal model of larch plantation stock volume was found, and the spatial distribution of Larix principis-rupprechtii plantation stock volume at different ages in the study area was obtained. The results showed that the prediction accuracy of regression model with the normalized backscattering coefficient and the index of stock volume of Larix principis-rupprechtii plantation of ALOS PALSAR radar data was the highest among the four regression models (R2 = 0.67, relative RMSE = 26.78%, P <0.01). According to the model, the total volume of Larix principis-rupprechtii plantation in the study area was 4.7 × 106m3 and its corresponding spatial distribution map. When the unit volume reached 250 ~ 300 m3 · hm- 2, the normalized backscattering coefficients of the ALOS PALSAR radar data were insensitive to the change of stock volume and reached saturation. For the forest age <20 a, 20-30 a, 30-40 a and> 40 a The stock volume of Larix principis-rupprechtii plantation was 78, 97, 136 and 127 m3 · hm-2, respectively. This study presents a method to accurately estimate the stock volume of plantations based on radar images.