Obligis E. auteur aut IRD Rahmani A. auteur aut IRD Eymard Laurence auteur aut IRD Labroue S. auteur aut IRD Bronner E. auteur aut IRD text journalArticle eng Algorithms proposed to retrieve water vapor are usually developed using statistical regression. Recent results have shown that this global statistical approach creates systematic geographical errors implying systematic biases. In the particular case of the wet tropospheric correction for altimetry missions, these biases directly distort the sea level estimation. After an in-depth analysis of the retrieval errors, we propose a new approach that significantly improves the quality of the retrieval algorithm. specialized Microwave radiometry water vapor retrieval wet tropospheric correction 126 021 47 9 3057-3064 2009 0196-2892 https://www.documentation.ird.fr/hor/PAR00004047 10.1109/tgrs.2009.2020433 0196-2892 https://www.documentation.ird.fr/hor/PAR00004047 IRD - Base Horizon / Pleins textes 2009-10-07 2017-08-23 PAR00004047 fre