About this page

Background to the production of weather forecasts, particularly the Shipping Forecast, that may give users a better understanding of that service.

On this page -


Sailors have many misconceptions and misunderstandings about the weather, weather forecasting and which of the many forecasts should be used. Another page suggests the most appropriate forecasts for different kinds of sailing.

The “flagship” is the UK shipping forecast with which we all have a love-hate relationship. It is often criticised for its “accuracy” or lack of it without any thought as to how accuracy is defined for a forecast –

  1. Covering a 24 hour period,
  2. Describing weather over a 1M square mile area,
  3. Written in no more than 330 words including the preamble and gale warning summary.

Background to the BBC Shipping Forecast

Meteorological observations from fixed observing stations, merchant ships, fixed buoys, aircraft and weather balloons are collected on a world-wide basis. For this purpose, the two main observing times are 00 and 12 UTC. Slightly less data are collected at 06 and 18 UTC. Observations at these four times are exchanged globally. Data exchange, between western Europe countries occurs every hour. In addition, there are data at irregular times from peripatetic sources such as aircraft, drifting buoys and satellites. All these data are used by the numerical weather prediction models.

The Met Office computer model uses 00 and 12 UTC data to provide forecasts out to 6 days ahead. The 06 and 18 UTC data provide forecasts only to 48 hours. The difference is because many stations launch radio-sonde balloons at 00 and 12 but far fewer at 06 and 18 UTC. Every effort is made to ensure that all data, regardless of time and type, are used as effectively as possible. The data are analysed using a 4-D "best fit" approach (in space and time).

Initial output from the computations using global data are available about 3-3½ hours after data time. When working in BST, this allows a very short time for the forecaster to complete the shipping forecast. To meet the 0520 and 1754 LT schedules on BBC Radio 4, he has to have the script, already in draft, to be finalised very quickly when he sees the latest model output. The BBC have tried to bring the 1754 forecast forward by an hour but the Met Office said that that was simply not possible if the most up do date information was to be used.

Approximate time scales for the production of the basic shipping forecasts are shown in the following table:

Data Time (UTC)

Text Leaves Met Office (LT)

Broadcast Time (LT)













NOTE These timings will probably have changed a little but the table will still give some idea of the timescales involved.

In addition to the predictions from the computer models, the forecasters have more recent data from the British Isles, the near Continent, data buoys, oil rigs, light vessels and islands. These are all used to fine tune the model output and are scrutinised as carefully as time allows, particularly with regard to the need for gale warnings.

Why are forecasts sometimes so vague?

First and foremost the weather itself is so indefinite; it is a truism to say that the weather does not know itself to within one Beaufort force. On Channel or North Sea crossings, for example, we can experience variability in the wind about which the forecaster knows very little. As an example, a westerly F4 up the Channel will occur with a 4 hPa difference in pressure between the south coast of England and northern France, say 1000 hPa over Southampton and 1004 over St Malo. A force 5 would occur with 999 hPa over Southampton - a pressure difference of 0.1% - and 1004 over St Malo. Relatively small errors in a forecast can make all he difference between a nice force 4, an exciting force 5/6 and a nail biting force 7/8 with gusts to 9!

Secondly, marine forecasts cover large sea areas some with complicated stretches of coast. The wind can vary from force 3 outside the Dart to force 6 off Bolt Head, in about 15 miles and 3 hours. Imagine the length and complexity of the forecast if all such detail was included for even 12 hours ahead from "Lands End to Lyme Regis including the Isles of Scilly".

It does not help anyone if the forecaster tries to be more definite than he really can be. The forecaster does not like using such words as "perhaps", "possibly", "locally" but it is all too often necessary if he is to convey the possibilities in a sensible manner, achievable in a script that is not so long that it becomes meaningless.

Another truism is that there is nothing really definite in the forecast except for the date. Even the date is subject to a certain amount of theological discussion!



There are various sources of GMDSS weather information available around the UK and Western Europe. These come originally from the various National Met Services and are free at the point of delivery. Most other forecasts will be entirely automated. By and large these will be very similar to the GNDSS product but will be less reliable as indicators of bad weather – wind strength and visibility etc,

Find which services meet your need and which you can obtain the easiest. When listening to a forecast, do not rely on memory. Make brief notes and transcribe them immediately afterwards, if only to save arguments with your crew!

Good luck and fair winds. If not fair winds then the least that I can wish you is foul weather well forecast.

A version of this page was originally published in Cruising, the quarterly magazine of the Cruising Association.

Return to Home Page