Using Principal Component Analysis in Radiative Transfer Calculations

Atmospheric optical properties depend on the chemical distribution in the atmosphere and the wavelengths of the solar and terrestrial radiation that interact with it. Through scattering, absorption and emission of radiation within and between various atmospheric layers, these optical properties affect the transfer of radiation through the atmosphere. Radiative transfer determines atmospheric cooling/heating (that in turn drives atmospheric circulation) and provides the basis for satellite remote sensing. However, accurate radiative transfer calculations are computationally expensive. To accelerate these calculations, we investigate a new way, employing Principal Component Analysis, to determine the dominant modes of variability in the vertical and spectral structure of the atmospheric optical properties.