Comments on "A statistical determination of the random observational errors present in voluntary observing ships' meteorological reports''

The accuracy achievable for surface ship pressure reports and the sources o f error in them are discussed. In The Met Office numerical weather predicti on (NWP) system the error is estimated as 1 hPa, whereas Kent et al. calcul ated a figure of about 2.3 hPa. There are a number of reasons for this disc repancy, but the main one is that quality control using short-range forecas ts is much better at detecting gross errors in the reports than are climato logical checks. Statistics from the Met Office NWP system are presented for comparison of pressure and other atmospheric variables. Other variables sh ow much less discrepancy than pressure because, relative to observation err or, they have a smaller climatological range. Both sets of statistics show larger air temperature errors at high latitudes/low temperatures.