Sometimes the extent to which anyone is right or wrong depends on who the umpire is. As the 16th-century satirist John Harington put it: Treason doth never prosper: what's the reason? For if it prosper, none dares call it treason.
Likewise, in the spiritual world, the Roman Church has solved the problem of the possibility of error by declaring its head to be infallible. The doctrine of the infallibility of the Pope, although openly contested at the time by many good and very learned bishops, was proclaimed 130 years ago today on July 17th, 1870. Once Roma lucuta erat , that was that; ipso facto, no Pope has got it wrong since then.
But no such ploys are available to those ordained to forecast weather. Even when making a solemn proclamation to the world on matters solely concerned with meteorology or climate, they are not deemed to be immune from error. Sometimes a weather forecast can be wrong.
A forecast, however, is very seldom wrong in all respects; likewise, it is very rarely absolutely right. The message consists of information on a wide range of different elements like wind, temperature, rain, and visibility.
A forecast thought by one individual to have been good, because it was correct in the matters which concerned him, may well have been a failure to another who has different interests. The sailor, for example, may care little about the frost which bothers horticulturists, but it may be a matter of life and death to get a timely warning of a storm.
One simple way of assessing forecast accuracy is to concentrate on one weather element at a time. Forecasts of rain, for example, can be evaluated on a "yes/no" basis, depending on the occurrence or otherwise of rain at a particular time and place. This is not entirely satisfactory, since it ignores the heaviness or persistence of the rain, but it is some indication of performance nonetheless and gives a success rate of about 85 per cent.
Subjective methods of assessment are also sometimes used. A number of people can be asked to monitor the weather forecast over a period, to note carefully the actual weather that occurs, and then to rate the forecast, say, on a scale from one to five. Such exercises generally put the number of "acceptably good" forecasts at somewhere between 80 and 90 per cent. Whatever the method used, all assessments indicate that the commonly perceived improvement in the quality of weather forecasts in recent years is not illusory. Meteorologists are far from being infallible, but they are less prone to error than they used to be.