- Using the Web Site
- Review of Available Wave Data Sources
- Instrumental Data
- Hindcast Data
- Visual Data
- The Source Data
- The Marine Databank
- The NMIMET Method of Analysis
- Wind Statistics
- Scatter Tables (Wave Height and Period)
- Extreme Wave Height Estimation
- Persistence of Storms and Calms
- Reliability and Validation
- European Database
The Marine Databank
The Meteorological Office Main Marine Databank  or  is a computerized archive containing more than 55 million marine meteorological observations from the period 1854 to 1984. As the observations have all been made by ships of passage, the databank contains meteorological data for all sea areas of the world; it is updated regularly to incorporate the most recent data and ships send in their observations on a voluntary basis. For many sea areas the VOF observations are the only data available on marine weather conditions.
The Meteorological Office has always collected and archived marine data from British ships and, under the provisions of WMO Resolution 35 (which nominated eight countries, each of which was to be responsible for the collection and archiving of marine data from a nominated area [Figure 1], now holds a complete archive of marine data for the North Atlantic from 1960 to date as part of the Marine Databank. Exchange agreements with the United States, Netherlands, the Federal Republic of Germany, Hong Kong and Japan have meant that all available data for their respective areas of responsibility for the period 1961 to 1983 have been added to the Marine Databank and, in addition, American "historic" data (originating before 1965) have been merged to give cover from 1854 to date.
All observations in this Marine Databank have been quality controlled, as this is essential if the data are to be used with confidence. Obviously it would be very time-consuming to scrutinize each observation manually; therefore a computerized quality-control system was developed which was designed to process historic information as well as more recent data and those from both British and foreign ships. For observations made by British ships since 1960 there has been an initial brief scrutiny of the logbook, in order to identify obvious coding errors, followed by a two-stage automatic checking process.
The first stage of computerized checking is intended to identify keying errors or other obvious mistakes, and mainly involves fairly crude range checks. The second stage looks at the internal consistency of each observation and performs more detailed range checking.
At both stages, suspect data are either rejected or manually amended and, during the second stage, details of any amendments made are retained so that the original data can be restored if necessary.
Observations from foreign ships are subject to the same automatic quality control procedures as British ships, but the meteorological logbooks are not available for scrutiny. Historic data (i.e. originating before 1961) have also been subjected to the two-stage automatic checking process, but with a minimal amount of manual intervention. No attempt has been made to carry out a real quality control at this stage (based on comparisons with surrounding observations), because the locations from which observations are available are constantly changing as the VOF ships move between ports.
Although, ideally, the ships' observations would be randomly distributed over all sea areas of the world and over the years 1854-1984, this is not the case in practice. The density of observations is obviously much greater along major shipping routes (for example, the English Channel) than in other ocean areas (for example, the South Pacific), and there are many more observations from more recent years. Wind conditions are reported more frequently than wave conditions so that there are about three times as many wind observations as wave observations in the Marine Databank.