Verification of cloud motion winds

Abstract Effects of recent method changes to the METEOSAT wind extraction scheme are reviewed on the basis of monthly verification data of cloud motion winds with collocated radiosonde wind measurements. Their impact is estimated in quantitative terms using the empirical relationship between speed bias and reference wind speed as a tool to make the resulting performance changes visible. An equivalent relationship is shown for the RMS vector difference. Improvements are found for winds at all height levels, in particular at high level.