In 1985, the WMO Commission for Basic Systems (CBS) agreed that there was a need for global NWP centres to monitor the quality of observations available on the GTS and to exchange monthly lists of stations providing seemingly erroneous data. In 1988 three lead centres were nominated which would have a co-ordinating role of producing, at six-monthly intervals, consolidated lists of suspect stations for given data types, together with information on the nature of their errors. The Met Office was allocated the role as lead centre for marine surface observations, encompassing reports from ships, drifting buoys, moored buoys and other fixed marine platforms.

For each platform identified as suspect, values are supplied for the number of observations received at the Met Office, the number of these observations with gross errors, the observations' mean differences from the background values produced by the numerical data assimilation system and the standard deviation of these differences.

The Biannual report also maintains statistical records of observation totals over the period of it's publication. Figure 1 shows the number of marine pressure observations received at the Met Office since publication of the first Biannual report in 1989 to the latest report for July to December 2011. Over this period, the total number of pressure observations has risen more than 8-fold from about 600,000 to around 5 million. Reports from drifting buoys and, to a lesser extent, automatic ships (including moored buoys) are largely responsible for this significant increase, whereas the number of manual reports from ships has remained relatively static.

Further details of the monitoring methods employed, as well as statistical trends in the quality of surface marine observations and suspect lists can be found in the report for the latest period which may be downloaded in pdf format from the links below:

Biannual Report on the Quality of Marine Surface Observations - July to December 2011 (~1.0MB)

Figures One to Fourteen (~0.8MB)

Pressure Suspect Identifier Time-series Plots (~1.0MB)
Wind Speed Suspect Identifier Time-series Plots (~0.2MB)
Wind Direction Suspect Identifier Time-series Plots (~1.0MB)
SST Suspect Identifier Time-series Plots (~0.7MB)

Previous Report