The objective of this project is to understand and diagnose the values of Stokes 3 and 4 observed by SMOS. For this purpose, SMOS+ Polarimetry will do the following:

1- To make the difference between the geophysical signal and various biases coming from the instrument or the reconstruction process;

2- To understand the possible link between the biases observed for the four polarizations;

3- To justify the implementation of a correction either at Level 1a or, at Level 1c. These two approaches will be investigated in detail;

4- To apply these corrections to a representative SMOS dataset in order to assess the validity limits of these corrections.

In line with the previous section, all the polarization components will be taken into account for the diagnosis and the development of a correction; the final objective being to propose a correction to be implemented in the operational chain.

From the test results performed during this study, we mean to determine whether it is possible or not to correct biases on TBs. Otherwise, the bias impact on the estimator of salinity will be estimated. If this impact is too significant, then it is clear that St3 and St4 polarizations will not have to be used in the L2-OS processing. Note that for the moment, these polarizations are used in the operational chain after an additional OTT correction. However, the estimated Total Electronic Content (TEC) has no physical meaning.

At the end of the commissioning phase, the choice to operate SMOS rather in full polarisation mode than in dual polarisation mode was justified by the exploratory scientific mission of the SMOS programme. The fact that SMOS can collect data in full pol mode is of several interests:

A) St3 and St4 are expected to be weak (if one excludes the Faraday rotation and the rotation from ground reference frame to antenna frame). It will be interesting to see how the observed signal is in line with the expected signal.

B) Since St4 must be weak, perhaps a signal significantly different from 0 allows helping the diagnosis of instrumental biases. This could have an effect on the understanding of biases affecting X and Y polarizations and thus also allow to better correct them.

Until now, the follow-up of TBs does not allow to clearly explain the St3 and St4 observed signals. This study will be undertaken in this context and aim at bringing answers to these questions.