A numerical weather prediction (NWP) model is a computational system that divides the atmosphere into a three-dimensional grid and solves the equations of fluid dynamics, thermodynamics, and radiation for each cell at successive time steps, simulating the future evolution of the atmospheric state. It forms the foundation of modern weather forecasting, complementing human interpretation of synoptic charts. The main global models are: GFS (Global Forecast System, NOAA—~13 km resolution), ECMWF/IFS (European Centre, considered the most accurate—~9 km resolution), GEM (Canada), and ICON (DWD Germany). Higher-resolution limited-area models also exist: HARMONIE-AROME (AEMET, 2.5 km for Spain), AROME (Météo-France, 1.3 km), and HRRR (NOAA, 3 km for the USA). Resolution determines which phenomena can be resolved: at 1–3 km individual thunderstorms are resolved; at 10+ km only larger-scale systems are represented. Ensemble forecasting runs multiple simulations with slightly perturbed initial conditions to quantify uncertainty. Reliability decreases with forecast range: 1–3 day predictions are highly accurate, 5–7 days are useful, and beyond 10 days only general trends are available. Satellites, radiosondes, and radars provide the input data (observations) that feed these models.