A Forest Leaf Area Index Inversion Scheme Using Polarimetric SAR

A novel forest leaf area index (LAI) inversion scheme using polarimetric synthetic aperture radar (PolSAR) images is proposed. The advantage of inversion using synthetic aperture radar (SAR) images over conventional optical methods is the ability to perform all-weather and all-time forest observation, especially in the tropical rainforest area. In this article, leaves were assumed to be isotropic media, while branches were assumed to be anisotropic with a certain orientation. We first investigated the specific extinction coefficients of propagation through leaves and branches, respectively. Then, a method was employed to separate the attenuation of leaves and branches. Subsequently, an inversion scheme model for LAI inversion using PolSAR images was designed based on the linear relationship between the LAI, and the products of forest height and extinction coefficient. Finally, an experiment was performed using L-band Satélite Argentino de Observación COn Microondas (SAOCOM) spaceborne SAR image pairs. The forest height and extinction coefficients according to H and V polarizations were retrieved using the oriented volume over ground (OVoG) model. The linear relation between the LAI, and the products of forest height and extinction coefficient were verified using the data measured in situ, and the coefficient of determination (
${R} ^{2}$
) was 0.64. Finally, the temporal decorrelation and the extinction component caused by the oriented elements were analyzed.