应用双频合成孔径雷达(SAR)数据和干涉水云模型估算森林生物量

Estimating Forest Biomass Using Dual-Frequency Synthetic Aperture Radar(SAR)Data and an Interferometric Water Cloud Model

以河北省承德市塞罕坝机械林场为研究区域,应用哨兵1号(Sentinel-1A)C波段双极化数据和大地2号L波段全极化数据分别建立干涉水云模型(IWCM)估算森林地上生物量。结果表明:大地2号(ALOS-2)L波段全极化数据体散射比例为72%,二面角散射比例为4%;L波段HV极化和VH极化的估算结果较好,HV极化估算结果的决定系数(R^(2))为0.737,均方根误差(R_(MSE))为28.88 t/hm^(2);VH极化估算结果的决定系数(R^(2))为0.743,均方根误差(R_(MSE))为27.76 t/hm^(2);C+L波段的组合方式反演结果最优,决定系数(R^(2))为0.863,均方根误差(R_(MSE))为18.92 t/hm^(2)。采用干涉水云模型估算森林生物量,L波段合成孔径雷达(SAR)数据估算精度高于C波段估算精度,L波段SAR数据更适合进行森林生物量的估算;C波段和L波段SAR融合对森林生物量的估算精度显著提高。

In the study area of Saihanba Mechanized Forest Farm in Chengde City,Hebei Province,dual-polarization C-band data from Sentinel-1A and fully polarimetric L-band data from ALOS-2 were used to establish an Interferometric Water Cloud Model(IWCM)for estimating forest aboveground biomass.The results showed that the L-band fully polarimetric data from ALOS-2 exhibited a volume scattering ratio of 72% and a dihedral angle scattering ratio of 4%.The estimation results of L-band HV polarization and VH polarization are good,with a coefficient of determination(R^(2))of 0.737 and root mean square error(R_(MSE))of 28.88 t/hm^(2) for HV polarization,and R^(2) of 0.743 and R_(MSE) of 27.76 t/hm^(2) for VH polarization.The combination of C and L bands yielded the optimal inversion results with an R^(2) of 0.863 and R_(MSE) of 18.92 t/hm^(2).Employing the IWCM for biomass estimation,the L-band SAR data demonstrated higher accuracy than the C-band SAR data,indicating L-band SAR data as more suitable for forest biomass estimation.Fusion of C-band and L-band SAR significantly improved biomass estimation accuracy.