Why can we not predict
what the weather will be like?
Why can we not predict what the weather will be like?
Not a long time ago, weather forecasts were very unreliable. Today the reliability of the forecast for the following day is around 95 percent, and good computer models can give a reliable forecast for one week ahead. However, as we look further in time, the reliability decreases and predicting weather conditions for a long period is still a ''mission impossible''.
The problem with computer models which ought to predict weather conditions for longer periods appears because it is a typically chaotic system, which means that even the slightest difference in starting conditions lead to different final results. The meteorologist Edward Lorenz pointed to this in 1961, when he was trying to make a long term forecast and discovered a so called "Butterfly Effect". To put it shortly, by waving its wings in Brasil, a butterfly can cause a tornado in Texas.

Contemporary climatologists and meteorologists try to make seasonal forecast; that are actually experimental forecasts trying to show whether the coming period will be warmer or colder than average.