LAUSR

Rigorous electromagnetic computations required for the calculation of high-resolution monochromatic bulk integral optical properties of irregular atmospheric particles are onerous in memory and in time requirements. Here, it is shown that from a set of 145 monochromatic bulk integral ice optical properties, it is possible to reduce the set to eight hinge wavelengths by using the method of principal component analysis (PCA) regression. From the eight hinge wavelengths, the full set can be reconstructed to within root mean square errors of ≪1%. To obtain optimal reconstruction, the training set must cover as wide a range of parameter space as possible. Rigorous electromagnetic methods can now be routinely applied to represent accurately the integral optical properties of atmospheric particles in climate models.