The First Forecasts

Predicting the weather became possible not when meteorology improved but when the telegraph arrived. A storm that reaches a coast has usually spent a day crossing the sea, and a message that travels faster than the storm turns an observation in one place into a warning in another. The science of the eighteen fifties could not say why the storm existed; it could say where it had been that morning, and for a fishing fleet that was the useful part.

The early services were therefore built on reports rather than on theory, and the arrangement had an unexpected cost. Because the forecasts were plainly useful, they were also plainly wrong from time to time, and each failure was public in a way that a mistaken laboratory result never is. Several early forecasters were attacked as charlatans and one national service was suspended for a decade. The episode is a reminder that a prediction is judged by a different standard from an explanation, and that a discipline which begins to predict before it can explain buys its usefulness with its reputation.