GPD Wet Tropospheric Correction for all ESA and Reference Altimetric Missions
Fernandes, M. Joana¹; Lázaro, Clara¹; Nunes, Alexandra L.²; Pires, Nelson¹; Ablain, Michaël³; Benveniste, Jérôme^4
1Univ. Porto; CIIMAR, PORTUGAL; ²ISEP; CIIMAR, PORTUGAL; ³CLS, FRANCE; ⁴ESA, ITALY

The main purpose of the ESA Climate Change Initiative (CCI) Sea Level project is to produce and validate the Sea Level Essential Climate Variable (ECV) product. For this purpose, the best algorithms for climate applications were developed, tested and selected.
As part of this effort, the GNSS-derived path delay (GPD) algorithm, for computing the path delay in the altimeter measurements due to the presence of water vapour in the atmosphere, developed by the University of Porto (UPorto) was selected as the best candidate for use in the generation of the final sea level ECV. In the sequel, the algorithm which has initially been designed as a coastal correction for Envisat in the scope of the ESA project COASTALT, has been extended to open ocean and tuned to all reference (TOPEX/Poseidon, Jason-1, Jason-2) and ESA (ERS-1, ERS-2 and Envisat) missions.
This paper gives an overview of the GPD implementation for all the six altimetric missions with emphasis on the impacts on the sea level ECV.
The basis of the algorithm is the data combination, by objective analysis of three main wet tropospheric correction data types: valid measurements from the microwave radiometer (MWR) on board the altimetric mission, wet path delays derived from Global Navigation Satellite Systems (GNSS) coastal stations, and those derived from the European Centre for Medium-range Weather Forecasts (ECMWF) ReAnalysis (ERA) Interim model. For each point along the satellite track, according to pre-defined criteria, the algorithm decides if a new estimate is to be computed or if the original MWR-based correction is kept unchanged.
For each mission, the state-of-the art correction from the onboard MWR has been used and the algorithm was tuned to allow a proper detection of the points at which the wet tropospheric correction has to be estimated (either due to land, ice or rain contamination, or to instrument malfunction). Prior to the computations the three different data sets were inter-calibrated using ERA Interim as the long-term reference.
The GPD products have been validated by comparison with other wet tropospheric corrections, such as the so-called composite correction, and by sea level anomaly (SLA) variance analysis at crossover points and function of distance from the coast. The highlights of these results are presented with emphasis on the impacts of the correction in the building of the 20-year long altimeter record for climate studies.