An analysis of the factors influencing the modeling accuracy of the invariant imbedding T-Matrix method and the optimal design of the parameter settings for particles with different geometrical and optical properties

Abstract

The Invariant Imbedding (IIM) T-matrix method has been recognized as one of the most promising method to derive exact solution of the light scattering properties of the nonspherical atmospheric particles after the significant improvement by Bi L. and Yang P. Compared with DDA (Discrete Dipole Approximation) and the time domain methods, it can calculate the scattering properties of the random-oriented particles analytically once the T-matrix of the particle is obtained. In order to better understand the modeling capability of this model and optimize the parameter settings for the particles with different geometrical and optical properties, the factors influencing the modeling accuracy are systematically analyzed, which includes the discretization schemes of the radial integral, the discrete point numbers of the radial, zenith and azimuthal integrals. The results show that, the calculation accuracy of Gaussian quadrature technique is higher than that of the trapezoidal rule in the radial discretization. For particles with large size parameter or refractive index, the density of the discrete points (both for the radial discretization and the mesh generation of the zenith and azimuth angles) should be increased correspondingly to ensure the calculation accuracy. Finer discrete grids are also required for particles with extreme shapes (e.g., particles with large aspect ratios). Based on the analysis above, the method to determine the parameters of scattering simulation is further proposed for particles with different geometrical and optical properties, by which the discretization of the computational domain can be optimized once the refractive index and geometry of the particle are known. The method is also validated for cylinders and hexagonal prisms, the results shows that the optimization scheme can be applicable to particles with different shapes.