Research On Retrieval Algorithm Of Forest Leaf Area Index Based On SAR Scattering Process

Leaf area index(LAI)is one of the key parameters to characterize sparseness of leaf,the structure and growth status of vegetation canopy,which can present the photosynthesis,respiration,transpiration and the carbon circle,etc.So,LAI is an important input parameter to study local and global terrestrial ecosystem-atmosphere interactions,vegetation dynamics and surface models.In the study of the global forest ecosystem,LAI has been considered as an important index for describing the status of global forest and the cycle of matter.Therefore,it is crucial to obtain forest LAI quickly,objectively and accurately.Remote sensing is the only technical means for large-scale forest LAI inversion,and the all-weather observation capability of active microwave remote sensing can overcome the limitation of optical remote sensing to obtain ideal data in tropical rainforest areas with complex climate,and can provide a reliable observation means for global forest observation.Among,the Polarimetric interferometric SAR(Pol In SAR)technique combines the advantages of polarization and interferometry techniques and becomes an indispensable technique in forest parameter inversion.The active microwave signal emitted by SAR has a strong penetrating ability,and the leaf is the main scatterer in the forest volume layer.The SAR signal propagates in the forest volume layer through the process of reflection,scattering and attenuation,which is related to the information of the number density,size and direction parameters of the scatterers in the volume layer,and these parameters also reflect the size of the LAI.At the same time,different wavelengths,incident angles and polarization methods will present different attenuation and scattering characteristics,and their rich information makes SAR techniques have great potential for application in extracting leaf area indices.In this paper,based on the simulation analysis of the attenuation characteristics of electromagnetic waves in the forest canopy and the study of the remote sensing mechanism of Polarimetric interferometric SAR,a study on the inversion of the forest leaf area index using Polarimetric interferometric SAR technique is carried out,and the main work is as follows.(1)Based on the three-dimensional structure of the forest,the lossy dielectric discs and finite-length dielectric cylinders with the same scattering characteristics as the forest scatterers are selected to simulate the leaves and branches,and the internal field of the leaves and branches are estimated using the generalized Rayleigh-Gans(GRG)approximation model and the infinite-length lossy dielectric cylinder scattering model to simulate and analyze the electromagnetic wave attenuation characteristics of different scatterers in the forest volume layer.The results indicate that the attenuation contributions of leaves in the forest volume layer are more significant in the high frequency domain,especially at 6 GHz(C-band),where the attenuation contributions of broadleaves to horizontal and vertical polarization waves reach 90.7%and 79.6%,respectively,and those of needles reach 69%and 77.2%,respectively,both of which are more than half of the total attenuation of the entire forest canopy.At last,the variations in the attenuation of the polarization waves with the permittivity of scatterers and leaf shape were also discussed.(2)A linear estimation model of the forest LAI based on the polarization scattering mechanism is proposed by combining the distribution characteristics of leaves and branches in the forest.The attenuation contributions of the leaf and branch scatterer sets are firstly separated from the coupled polarimetric interferometric complex coherence coefficients,and then a linear physical relationship between the leaf area and the product of tree height and attenuation coefficients is established based on the definition of the forest LAI.The model relates the forest LAI to the statistical properties of leaves and branches parameters in the forest volume layer from the perspective of polarization scattering attenuation,providing effective information for inversion of forest leaf area using radar data.(3)A forest leaf area index inversion algorithm based on Polarimetric interferometric SAR technique is proposed.The algorithm uses the OVoG model that fits the distribution characteristics of forest scatterers,takes the L-band SAOCOM fully polarization SAR images as the data source,inverts the information of tree height and attenuation coefficients of horizontal and vertical polarization waves of the forest in the study area,obtains the product value between tree height and attenuation coefficients,and finally inverts the leaf area parameter K·LAI.then the accuracy of the algorithm is verified by using the actual measured LAI data.The results show that the R~2 between the actual measured LAI and the inverse-derived K·LAI is 0.65,and the two have good correlation,which proves the feasibility of the inverse forest LAI scheme based on Polarimetric interferometric SAR technique from theoretical and experimental perspectives.Finally,the effects of time de-correlation,attenuation components of directional elements and K values on the results are analyzed.Figure[24]Table[2]Reference[109]...