基于极化定向角补偿的思茅松林地上生物量反演

目的 极化合成孔径雷达在森林遥感监测中得到了广泛的应用。由于法拉第旋转和地物结构特性,电磁波极化定向角发生偏移,导致散射特征在机理上存在模糊性。本研究主要分析极化定向角偏移对体散射分量和地上生物量反演的影响。方法 以ALOS PALSAR全极化星载合成孔径雷达(SAR)数据为数据源,基于L波段散射特征,考虑地面与树干之间的二面角散射贡献,研究提出了一种扩展极化水云模型;基于Yamaguchi四分量分解参数和扩展极化水云模型估测思茅松林地上生物量。结果 通过酉变换来补偿极化定向角偏移后,体散射分量高估得到修正,极化定向角补偿后的体散射与实测地上生物量的回归模型较未补偿前效果更好(决定系数R²从0.214提升到0.332)。采用Yamaguchi四分量和扩展极化水云模型的地上生物量估测值和实测值有较强的相关性(R²= 0.644)和较低的均方根误差(23.11 t/hm²)。结论 SAR数据在极化分解前应进行极化定向角补偿,以减少体散射高估和二面角散射低估的问题,提高地上生物量反演精度。半经验极化扩展水云模型具有很好的估测森林地上生物量的潜力。

Based polarization orientation angle compensation for Pinus kesiya var. langbianensis forest aboveground biomass estimation

【Objective】Polarimetric synthetic aperture radars have been widely used in forest remote sensing monitoring. Owing to Faraday rotation, the polarization orientation angle (POA) of the electromagnetic wave is displaced, leading to ambiguity in the scattering characteristics. In this study, the effects of polarization orientation angle compensation on the volume scattering component and aboveground biomass (AGB) retrieval were analyzed.【Method】 The influence of Faraday rotation on SAR data was analyzed using ALOS PALSAR full polarimetric SAR images as the data source. Based on the L-band scattering characteristics and considering the dihedral scattering contribution between the ground and the tree trunk, an extended polarization water cloud model (EPWCM) was proposed. Based on the Yamaguchi four-component decomposition parameters and field survey data, the aboveground biomass of Pinus kesiya var. langbianensis forest was estimated by EPWCM.【Result】 Through the unitary transformation of the coherence matrix to compensate for the polarization orientation angle deviation, the overestimation of the volume scattering component was corrected, and the regression with aboveground biomass was improved (R ² increased from 0.214 to 0.332). The estimated aboveground biomass had a strong correlation with the observed AGB (R² = 0.644) and a relatively high accuracy (RMSE as 23.11 t/hm²). 【Conclusion】 Before polarimetric decomposition, SAR data should be compensated for polarimetric orientation angle correction to reduce the ambiguity of scattering characteristics and increase the retrieval accuracy of AGB. The semi-empirical model has a good potential for estimating forest aboveground biomass.