Q1.
Why do weather forecasts sometimes go wrong?
Answer
Because weather depends on many quantities that keep changing together, and very tiny differences in today's conditions grow into completely different weather a few days later.
- The state of the atmosphere is set by temperature, pressure, humidity and wind — at every point, and at every height.
- Forecasting uses measurements fed into a model. But the measurements come from stations and satellites spaced kilometres apart, so what happens between them has to be filled in.
- A small error in that starting picture does not stay small. It doubles, doubles again, and after a few days it is as large as the weather itself.
error today ≈ 0.1 °C at one point
grows roughly by doubling every 1–2 days
after 8 days ≈ 0.1 °C × 2⁵ ≈ 3 °C — the size of a whole weather change
grows roughly by doubling every 1–2 days
after 8 days ≈ 0.1 °C × 2⁵ ≈ 3 °C — the size of a whole weather change
Why it happens: the atmosphere is not simply complicated, it is sensitive. The equations of air flow feed back on themselves — a slightly warmer patch rises a little faster, which pulls in more air, which shifts a cloud, which shifts where the sun heats the ground next. This is why forecasts are usually reliable for a few hours or even a few days, but far less certain further into the future. Note carefully what is not the reason: the physics is not wrong, and the model is not badly built. The limit comes from never being able to measure the starting state perfectly.
Did you know? Forecasters get around this by running the same model many times from slightly different starting conditions. If 45 of 50 runs give rain, they announce “90% chance of rain”. The spread of the runs is itself the honest measure of how far ahead the forecast can be trusted.