Speaker
Description
The processing of limb radiance observations is a difficult task due to the complexity of the forward model where several interfering species may interfere within multiple orders of scattering in the radiative transfer model. Furthermore, the observation geometry influences the observed radiance through the air masses, phase function and polarization.
In general, the inversion of limb radiances is performed by iterative methods due to the strong nonlinear character of the problem. The cost of the evaluation of the merit function is an important driver of the algorithm efficiency if many successive inversions have to be performed.
Here, we explore “direct” methods where a mapping is constructed between a large ensemble of limb radiance simulations and the set of atmospheric constituents used to produce them.
In the frame of the ALTIUS mission, we expect interferences between ozone and aerosol contributions depending on the atmospheric volcanic load. As an example of the power of direct methods, we address the issue of deriving aerosol properties and number density profiles.
