Auto-adaptive Tikhonov regularization of water vapor profiles: application to FORUM measurements

In this paper, we study the retrieval of water vapor profiles from simulated FORUM measurements. We show that the bias towards the a-priori introduced by the Optimal Estimation technique can be reduced by using larger errors for the a-priori. Reducing the strength of the a-priori may, however, cause unphysical oscillations in the resulting profiles because of the ill-conditioning of the retrieval problem. An a-posteriori regularization technique, the Iterative Variable Strength method, is thus applied to reduce the amplitude of the oscillations.