Information content analysis of above-cloud aerosol retrievals

Above-cloud aerosols could substantially impact Earth's energy budget, causing more or less radiation absorption in cloudy scenes than would be expected (for absorptive and reflective aerosols respectively). However retrievals regarding above-cloud aerosols are usually quite challenging to perform. It can be difficult to know a-priori which properties of the aerosol layer can be retrieved with any certainty. In this study we modelled a layer of above-cloud aerosols with varying physical properties and size distributions. We used an adding doubling radiative transfer code to determine the reflectance (both total and polarized) at the top of the atmosphere. Using this process, as well as instrument calibration information and meteorology, it is possible to determine which measurements are meaningful by using the Shannon information content of each measure as a metric. Additionally a singular value decomposition of the sensitivity matrix, generated based on the modelled aerosol layer and associated reflectances, revealed which aerosol micro-physical properties produce the largest effect on these measurements. This technique, while applied primarily to MODIS bands in this study, can easily be adapted to any instrument. This tool could be invaluable in both analysis of existing instrument data and in the design of new instruments by providing a quantitative method for determining which measurements are optimal and which parameters can be realistically retrieved.