What is NWP?
NWP - Numerical Weather Prediction
It refers to weather forecasting done by running computer models that simulate the atmosphere using physical equations (wind, temperature, humidity, pressure, etc.).
Meteorological agencies (like ECMWF, NOAA, Met Office) use NWP models to generate weather forecasts.
So, in short:
NWP = computer-based weather forecasting using physics-based models.
How we use it?
Our NWP sources
| Name | Description |
|---|---|
| MetNO | Norwegian Meteorological Institute (used in services like yr.no)<br> Description: MetNO provides weather forecasts that are often based on models such as ECMWF (European Centre for Medium-Range Weather Forecasts) or MEPS (MetCoOp Ensemble Prediction System).<br> Use case: Regional and medium-range forecasts, especially over Northern Europe. |
| ICON | Icosahedral Nonhydrostatic model Developed by: DWD (Deutscher Wetterdienst – German Weather Service) Description: A global and regional numerical weather prediction model. Known for its high resolution and modern grid structure (icosahedral grid). Use case: Both global and regional forecasting, used operationally by Germany and other partners. The global version runs four times daily (00, 06, 12, 18 UTC). Forecast horizons:+180 hours (7.5 days) for the 00 and 12 UTC runs. +120 hours (5 days) for the 06 and 18 UTC runs. Up to +78 h: hourly output Beyond +81 h: 3-hour intervals |
| GFS | Global Forecast System Description: A global numerical weather prediction model that provides forecasts up to 16 days ahead. It runs 4 times a day and is freely available. Use case: Widely used in global forecasting, ensemble models, and as input for local models. |
| HRRR | High-Resolution Rapid Refresh Description: HRRR is a high-resolution, short-range weather prediction model that provides forecasts every hour. It is convection-resolving, meaning it can explicitly simulate thunderstorms and other small-scale weather features. It covers the continental United States (CONUS) with a typical resolution of 3 km. Use case: Ideal for short-term forecasting (up to 18 hours), especially for rapidly changing weather events like thunderstorms, wind gusts, and localized precipitation. Commonly used in aviation, emergency response, and high-precision applications. |
| IFS | Integrated Forecasting System Description: IFS is a global numerical weather prediction model used operationally by ECMWF. It is known for its high accuracy in medium-range forecasts (up to 15 days ahead) and forms the basis for many other forecasting systems across Europe. The model uses advanced data assimilation techniques and runs both deterministic and ensemble versions. Use case: Global and medium-range forecasting, widely regarded as one of the most accurate models in the world. Used by many national weather services (e.g., through Copernicus, MetNO, etc.) and in climate and seasonal predictions. |
| NAM | North America Mesoscale. Description: NAM is a regional numerical weather prediction model focused on short- to medium-range forecasts over North America. It provides high-resolution forecasts (typically 12 km or finer in nested domains) and includes detailed physical parameterizations for terrain, land surface, and atmospheric processes. NAM incorporates real-time observational data and is updated four times daily. Use case: Used primarily for short-range weather forecasting in North America, including precipitation, temperature, and severe weather. It is commonly used by meteorologists for operational forecasts, aviation planning, and emergency weather response. |
| HRDPS | Environment and Climate Change Canada (ECCC) Description:HRDPS is a high-resolution, short-range numerical weather prediction model developed by the Canadian Meteorological Centre. It offers detailed forecasts over Canada, with a 2.5 km horizontal resolution, allowing it to resolve small-scale weather phenomena such as convective storms and localized wind events. HRDPS is part of the broader Canadian NWP system and is run multiple times per day, providing forecasts typically up to 48 hours ahead. Use case: well-suited for short-range, high-resolution forecasting needs, especially over complex terrain such as coastal, mountainous, and forested areas in Canada. HRDPS is used for forecasting severe weather events, wind and temperature profiles, and local-scale precipitation, and is valuable in sectors like energy, transportation, agriculture, and emergency management. |
| UKMO | UK Met Office (United Kingdom) Description:The UKMO model is the high-resolution numerical weather prediction system developed and operated by the UK Met Office. It provides global and regional forecasts, with configurations offering horizontal resolutions down to approximately 1.5 km over the UK domain. The model is based on the Unified Model (UM) framework, which supports both atmospheric and coupled (ocean–atmosphere) simulations. Forecasts are run several times per day and extend up to 7 days ahead for regional domains and 10 days or more for global runs. Use case:UKMO is widely used for short- to medium-range weather forecasting, particularly over the United Kingdom and Europe. Its high spatial resolution allows accurate representation of local-scale weather features, including convective storms, fog, wind, and precipitation events. The model supports applications in aviation, energy, marine operations, agriculture, and public weather services, as well as scientific research and climate-related studies. |