Abstract's details

Performances and benefits of a 1D-variational approach applied to the wet tropospheric correction for the Sentinel 3A and 3B topography missions

Bruno Picard (Fluctus SAS, France)


Ralf Bennartz (Vanderbilt University, US); Frank Fell (Informus, Germany); Estelle Obligis (Eumetsat, Germany)

Event: 2022 Ocean Surface Topography Science Team Meeting

Session: Instrument Processing: Propagation, Wind Speed and Sea State Bias

Presentation type: Type Oral

Contribution: PDF file


Statistical methods are usually used to provide estimations of the wet tropospheric correction (WTC), necessary to correct altimetry measurements for atmospheric path delays, using brightness temperatures measured at two or three low frequencies from a passive microwave radiometer on board the altimeter mission. Despite their overall accuracy over oceanic surfaces, uncertainties still remain in specific regions of complex atmospheric stratification.
Thus, there is still a need to improve the methods currently used by taking into account the frequency-dependent information content of the observations and the atmospheric and surface variations in the surroundings of the observations. Within this context, a one-dimensional variational (1D-Var) method has been developed, assimilating the relevant passive microwave observations to retrieve the WTC over ocean.
This physical approach has been applied to the whole Sentinel-3A and Sentinel-3B periods and validated over open ocean against an extensive set of references: the operational product, the GPD+ approach, the AIRWAVE water vapour product derived from SLSTR observations.
The exhaustive validation exercise allows to establish the 1D-var approach as a mature algorithm capable of retrieving the WTC at the level of performance of the operational product. The metric based on the variance of SSH at cross-overs shows that the performance of the 1D-var is even slightly better at global scale, resulting from a more contrasted geographical distribution where the 1D-var or the operational approach performs better, depending on specific geophysical conditions.
Moreover each value of the WTC retrieved with the 1D-var method also benefits from an associated uncertainty and a self-consistent quality flag that greatly facilitates the assessment of the product.

This study has been funded by Eumetsat.

Oral presentation show times:

Room Start Date End Date
Sala Grande Tue, Nov 01 2022,15:15 Tue, Nov 01 2022,15:30
Bruno Picard
Fluctus SAS