Datasets
The reliability diagram shows the reliability of the ECMWF seasonal forecast system (SEAS5) with ...
Interval/period: N/A
This chart shows the Relative Operating Characteristics (ROC) diagram for the three-month ...
Interval/period: N/A
This chart shows the spatial variation in the Relative Operating Characteristics (ROC) skill ...
Interval/period: N/A
The ECMWF seasonal forecasts (SEAS5) are produced every month with a 51-member ensemble at a ...
Interval/period: N/A
The 850 hPa level is usually just above the boundary layer and at this level the day-night variation in temperature is generally negligible...
Interval/period: N/A
This chart shows 7-day mean anomalies of 500hPa geopotential height from the ECMWF Sub-seasonal ...
Interval/period: N/A
500-1000 hPa thickness is a measure of the mean temperature of a column of the atmosphere between these pressure levels and can be used to distinguish between warm and cold air masses and...
Interval/period: N/A
This chart shows the spatial variation in the Anomaly Correlation Coefficient (ACC) for the ...
Interval/period: N/A
Aurora: a deep learning-based system developed by Microsoft. It is initialised with ECMWF analysis. Aurora operates at 0.1° resolution.
Interval/period: N/A
Aurora: a deep learning-based system developed by Microsoft. It is initialised with ECMWF analysis. Aurora operates at 0.1° resolution.
Interval/period: N/A
Aurora: a deep learning-based system developed by Microsoft. It is initialised with ECMWF analysis. Aurora operates at 0.1° resolution.
Interval/period: N/A
FourCastNet v2-small:a deep learning-based system developed by NVIDIA in collaboration with researchers at several US universities.It is initialised with ECMWF analysis. FourCastNet operates at 0.25° resolution.
Interval/period: N/A
FourCastNet v2-small:a deep learning-based system developed by NVIDIA in collaboration with researchers at several US universities.It is initialised with ECMWF analysis. FourCastNet operates at 0.25° resolution.
Interval/period: N/A
FourCastNet v2-small:a deep learning-based system developed by NVIDIA in collaboration with researchers at several US universities.It is initialised with ECMWF analysis. FourCastNet operates at 0.25° resolution.
Interval/period: N/A
GraphCast (Google DeepMind): a deep learning-based system developed by Google DeepMind.It is initialised with ECMWF analysis. GraphCast operates at 0.25° resolution.
Interval/period: N/A
Interval/period: Sun, 01/01/1860 - Mon, 12/31/2300
Interval/period: Sat, 01/01/1949 - Sat, 07/04/2026
Interval/period: Thu, 10/18/2018 - Wed, 07/31/2019
to contain all available observations from balloons ascents through the atmosphere.
The main geophysical variables are temperature, humidity, and wind speed and wind direction,
as functions of atmospheric pressure levels. The observation platforms include pilot balloons, radiosondes, and ozonesondes.
The main observation data sources are:
- the NOAA Integrated Global Radiosonde Archive (IGRA version 2),
- the NCAR Upper-Air Database (UADB), and
Interval/period: Tue, 01/01/1901 - Tue, 12/31/2024
These diagrams compare Continuous Ranked Probability Skill Scores (CRPSS) of ECMWF with ...
Interval/period: N/A
Interval/period: Sun, 01/01/1950 - Fri, 12/31/2100
Interval/period: Mon, 01/01/1979 - Mon, 06/30/2025
These charts show clustering of ENS members based on the 500 hPa height forecasts. ...
Interval/period: N/A
Interval/period: Sun, 01/01/1860 - Mon, 12/31/2300
Interval/period: Wed, 01/01/1800 - Fri, 12/31/2100