Weather forecasting

Weather forecasting predicts future atmospheric conditions by combining observations with physical models and statistical methods. Forecasts range from short-term local predictions to global outlooks spanning days or longer.

Numerical weather prediction: The use of computer models that solve equations governing the atmosphere to predict future weather. It turns atmospheric observations and physical laws into time-evolving forecasts.
Atmospheric dynamics: The study of atmospheric motion and the forces that produce it. The forecast's physical core is the changing motion of air.
Severe weather warning: An official alert that hazardous weather is occurring or expected soon in a specified area. Forecasts help authorities issue warnings before dangerous conditions arrive.
Climate prediction: The prediction of statistical properties of climate over seasons to decades rather than specific day-to-day weather. It addresses longer time horizons and distributions, not a sequence of particular weather events.
Forecast verification: The assessment of forecast quality by comparing predictions with observed outcomes. Verification reveals whether forecast systems are accurate, calibrated, and improving.
Data assimilation: The process of combining observations with a model state to estimate a system’s current conditions. It builds the best initial atmospheric state from measurements that vary in coverage and accuracy.
Weather observation: A measurement or report of atmospheric conditions at a particular place and time. Forecasts need observations to constrain the atmosphere's current state.
Nowcasting: Very short-range forecasting that emphasizes current observations and conditions over roughly the next few hours. It addresses immediate decisions where radar and rapidly updated data matter most.
Climate projection: A conditional simulation of possible future climate under specified emissions or other forcing scenarios. Unlike an operational weather forecast, it explores scenario-dependent long-term change.
Forecast uncertainty: The range of possible outcomes and confidence associated with a prediction. Communicating uncertainty prevents a single forecast from being mistaken for certainty.

Linked from 80 pages