Verification of Physical Optics Approximation by the Discrete Dipole Method in Calculations of Light Backscattering

In problems of atmospheric remote sensing, particularly in interpreting lidar signals from cirrus clouds, the accuracy of modeling light scattering by nonspherical randomly oriented ice particles plays a crucial role. Although the physical optics approximation is commonly used due to its computational efficiency, it does not always provide sufficient accuracy, especially when particle sizes are comparable to the wavelength of incident radiation. This introduces systematic errors into scattering matrix databases used for solving inverse problems. This study employs the discrete dipole approximation to verify the validity of the physical optics approximation. The primary focus is on comparing results obtained at wavelengths of 0.532 and 1.064 μm for particles with various shapes and sizes from two to eight wavelengths. It is shown that using the physical optics approximation in this case leads to relative errors of up to 20% in the determination of the linear depolarization ratio, as well as to underestimation of the backscattering intensity by half. The findings enable a more precise estimation of the applicability limits of the physical optics approximation and provide corrections for the existing scattering matrix databases. This, in turn, will improve the accuracy of lidar data interpretation and enhance the quality of microphysical retrievals for cirrus clouds.